RADIOACTIVITY.The Atomic Weight of Radium.Two new determinations of the atomic weight of radium have beenmade. In one1 a much larger quantity of radium was availablethan in previous determinations, the total weight of purified radiumchloride being 1.35 grams. It was observed that a part of thispreparation, which had been kept two years in a stoppered quartztube, had absorbed oxygen, for on heating to 300° in nitrogen nowater was evolved, but on fusion in hydrogen chloride it evolvedchlorine and water and lost 5 per cent. in weight. The surfwe ofthe quartz tube wits completely disintegrated by the action of theradiations, Quartz is an unsuitable material to use for thepreservation of radium preparations. By fractional crystallisationof the material from aqueous hydrogen chloride, the value of theatomic weight found rose rapidly to 225.95, and did not furtherincrease after numerous further fractionations. The atomic weightwas determined by the aid of the methods and apparatus employedby T.W. Richards by finding the ratios RaC12/AgC1 and RaCl,/Ag.Radium chloride, dried in nitrbgen below 200° and fused in aplatinum boat in hydrogen chloride at 900°, was the starting point,this compound being rendered less hygroscopic and more stable inair by fusion. The mean value obtained in both series of deter-minations was 225-95, individual estimations not differing from themean by more than 0.03. Both in methods of purification and ofestimation the work was in the main a repetition of Madame Curie’sinvestigation, in which the vaTue found was 226.34 on the presentvalues for silver and chlorine.The main points of difference werethe preliminary fusion of the salt in hydrogen chloride, the precipi-tation of the solution with silver in the cold, and the correction forthe solubility of the silver chloride by the nephelometer. Thesedifferences, taken in conjunction with the greater quantity ofmaterial employed, no doubt account for the slightly different valueobtained, so that the number 226.0 may be taken to represent theatomic weight of radium, within a very narrow margin of error,when purified and estimated by these methods.0. Honigschmid, Moimtsh., 1912, 33, 253 ; A., ii, 523.U 289 REP.-VOL IX290 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.I n the other determination 2 very small quantities of material,usually between 2 and 3 milligrams of radium bromide, wereactually employed, and the determination was carried out on amodified form of Steele and Grant's micro-balance. The finalmaterial was prepared by the fractional crystallisation of 330 milli-grams of about 70 per cent.radium bromide, derived from Cornishpitchblende, from water and aqueous hydrogen bromide. Themethod of determination was also new, the bromide being con-verted into chloride and vice versa by ignition in hydrogen chlorideor bromide respectively, which had the advantages that there wereno transferences of the material during the whole determinationand only gaseous reagents were employed. The result for theatomic weight was 226.36, a number which is almost identical withthat of Madame Curie. The result is interesting mainly on accountof the methods, both of purification and of estimation employed.On account of the small quantity of the material and the novelmethod of weighing adopted, it is, however, impossible to comparethis result with the others until the errors have been quantitativelyevaluated. The single result given, in which a value practicallyidentical with the international figure was obtained for the atomicweight of barium, using as the material 3.5 milligrams of thechloride, is not sufficient to allow of this being done.As isdiscussed in the following section, fresh determinations of theatomic weight of radium by methods, both of purification and ofestimation, differing as much as possible from those followed byMadame Curie, are called for before this constant can be consideredcompletely known.The International Radium Standard.The primary object of the International Radium Standards Com-mittee has been fulfilled.A standard containing 21.99 milligramsof radium chloride, prepared by Madame Curie by methods identi-cal with those which she employed in the determination of theatomic weight and sealed up in a thin glass tube, has been acceptedas the International Radium Standard, and is now preserved atthe Bureau International des poids et mesures, SBvres, near Paris.A careful comparison, by various y-ray methods of this preparation,with three others containing respectively 10.11, 31'7, and 40.43milligrams of radium, prepared by Honigschmid in the course ofhis atomic-weight determination, showed that all four preparationsagreed with one another within the limits of experimental error,and certainly within 1 part in 300.One of the Austrian prepara-R. Whytlaw-Gray and Sir William Ramsay, Proc. Roy. Soc., 1912, A, 86,270 ; A., ii, 413KADIOACTIVITY. 291tions is to be preserved as a secondary standard a t Vienna, and incountries requiring them the National Testing Laboratories are tobe provided with sub-standards measured against both the Inter-national and Vienna standards. The funds necessary for thepurchase of the radium, both for the International and for theBritish sub-standard, have been privately raised in this country,and it is expected that the British standard will shortly be preparedand handed over t o the Nationd Physical Laboratory, where regularprovision will be made for the official teating of radium prepara-tions.The results of the comparison made in Paris of Madame Curie’sand the Austrian preparations show that, provided a raw materialis employed, such as Joachimsthal pitchblende, containing no oronly negligible quantity of thorium, independent workers using thesame methods can reproduce at will radium standards agreeing withone another and with the accepted International standard.Ifthorium is present in appreciable quantity in the radioactivemineral, the presence of mesothorium in the radium extracted wlll,of course, entirely vitiate the results, as it will yield, in additionto the practically constant y-radiation of the radium, a continu-ously varying y-radiation of the same order of penetrating power,due to mesothorium and its products.The absolute purity of the preparations so obtained is anotherquestion.For the present it suffices that the degree of purity isdefinite and can be reproduced a t will, as this is the primaryrequirement of any standard.The absolute purity of radium chloride fractionated by crystal-lisation from barium chloride to a constant chemical equivalent hasalready been called in q ~ e s t i o n . ~ I n the fractional crystallisationof isomorphous salt mixtures, the end product of the process is thatof minimum (or maximum) solubility.It is not invariably thecase that this minimum solubility is possessed by one of the twopure components; it may belong t o mixed crystals possessing adefinite composition. I f Honigschmid’s value, 225.95, is the trueatomic weight of radium, that of lead, calculated from it by deducbing five times 3.99, the atomic weight of helium, would be 206-0,and of uranium, by adding three times 3.99, would be 237.92.Whereas, if it be supposed that the mixture of constant composi-tion obtained by fractional crystallisation still contains 1 per cent.of barium chloride, the atomic weight of radium would be 227.0,from which 207.05 and 238.97, numbers more in accord with theaccepted values, would be the calculated values for lead anduranium.Since, however, in Madame Curie’s final preparations3 W. Marckwald, Physikal. Zeilsch. 1912, 13, 732 ;. A , , ii, 323.u 292 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.an addition of 0.6 per cent. of barium chloride notably strengthenedthe barium lines in the spark spectrum, it is very difficult to believethat preparations prepared by these methods still contain as muchas 1 per cent. of barium chloride. The experimental investigationof the point now being made will be awaited with interest. If othersalts of radium and barium are fractionaIly crystallised, as, forexample, the bromide, the proportion between the two elements forthe product of constant composition should not be the same as forthe chloride if mixed crystals of minimum solubility exist.a-R ay s .Further progress has been made in the exact determination ofthe ranges of the a-rays, with a view to testing the Geiger-Nuttallrelation between the range of the u-ray and the period of the changein which it is expelled.* I n the uranium-radium series the difficultquestion as to the character of the very low-range u-rays of uraniumitself has been elucidated by a careful comparison by a specialmethod of the range of the u-rays of pure uranium oxide with thoseof ionium and polonium, under conditions such that the absorptionof the rays in the active film was identical in each case.The formof the curves connecting ionisation with range was, as is to beexpected for homogeneous a-rays, identical with ionium andpolonium, but the uranium curve was different, and could becompounded out of two such curves shifted relatively to oneanother by 4 mm.of range. This indicates that the two a-particles,known to be expelled from the uranium atom in its first change,differ in range by 4 mm. The separate ranges in air at 1 5 O weredetermined to be 2.9 and 2-5 cm.6Confirmation of this has been obtained by the redeterminationof the proportion of ionisation contributed by the u-rays of radiumand uranium respectively from a uranium mineral, in which theseelements are in equilibrium.6 Boltwood’s ratio of 45 : 100 7 isobviously at variance with the fact that the range of, and thereforethe ionisation contributed by, the single u-particle of radium isgreater than that of either of the a-particles of uranium.Theratio calculated, on the present exact data between range andionisation, assuming the ranges of the a-rays of uranium to be, it9recently found, 2.9 and 2.5 cm., and that of radium to bo 3.3 cm.,is 58:lOO. The experimental value, obtained by comparing theAnn. Report, 1911, 275.6 H. Geiger and J. M. Nuttall, Phil. Mag., 1912, [vi], 23, 439 ; A . , ii, 408.6 Stefan Meyer and F. l’aneth, Sitzztngaber. K. Akad. Wiss. IVien, 1912, 121,7 Ann. Repo~t, 1909, 269.Ha, 1404RADIOACTIVITY. 293a-rays from films of pure uranium oxide of negligible thickness withthose from the equivalent quantity of pure radium, was found tobe 57.3:100, which a t once confirms the direct determination ofthe ranges, and renders it improbable that uranium emits twoa-particles, one of range about one-half that of the other, as hadformerly been inferred.8The apparatus employed in the determination of the ranges ofthe a-rays of uranium, the construction of which involved consider-able experimental skill, has been utilised in a much-needed reviewof the ranges of the a-rays from the products of the thorium andactinium series.9 The collected final results for all three series arecontained in the following table:Ranges in cm.a tSubstance.Uraniuiii-1 .................U ran ium-lZ ..................Ioniuni ........................Radium .....................Radium emanation .........Rad iuni . A ..................Radium-C' ..................Radiuni-F ..................Th oriuin .....................1< adio t hori iini ...............Thorium -X ..................Thorium ernanation ......Thorium-A ..................Thorium-C ..................Thorium-C' ..................Radioactininm ............Actiniun-X .................Actinium emanation ......Actinium-A ..................Actinium-C ..................0".2.372.752-853-133.944 *506-573 *582.583'674-084.745 '404 *558.164 '364-175'406.165.121592.502.903 -003-304-164 7 55-943.772 -723.874'305 '005 *704.808'604-604-405 -706.505.40Initial velocityx lo9 cm./sec.1-471 -541 *561 *611 '741'822 '061-681-511 *701 *751-851.931.822'211 -801-771 *932 '021-89The Geiger-Nuttall relation has been considerably strengthenedand extended as the result.I n all three series the curves obtainedby plotting the logarithms of the radioactive constants and of theranges as ordinates and abscissze respectively are in each seriesessentially straight lines of similar slope. For products of similarperiod the range of the rays of the uranium products are least, thatof the actinium products greatest, the thorium products being inter-mediate. Although the differences between the three series aresmall, they are undoubtedly of very great interest. Any a-particleis experimentally indistinguishable from any other, so long thevelocities are adjusted to equality, it9 can quite easily be arrangedby causing the swifter first to pass through a suitable thickness ofan absorbing medium.A11 are radiant helium atoms. The law of8 Ann. &port, 1911, 274.Geiger and Nuttall, Phal. Mag., 1912, [vi], a, 647 ; A . , ii, 1022294 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.retardation of the velocity during passage through an absorbingmedium and the law of increase of ionising power with loss ofvelocity are identical for all a-particles, and are independent of thenature of the parent atoms from which they have originated. Theprecise initial velocity of expulsion appears to be a function of theperiod of transformation, which is the same for all the members ofone series, but is different for different disintegration series. Hencethere must be some peculiarity of atomic structure preserved intactin the successive members throughout a series of disintegrations,pertaining to that particular series, and distinguishing it from otherseries. I f these data are confirmed, they are therefore obviously ofprime importance in connexion with the evolution of the elements.The clear exceptions to the rule are the a-rays of thorium itselfand of radioactinium.For the former the great difficulty of pre-paring the material free from subsequent products and of the deter-mination, on account of the lowness of the range, may account forthe discrepancy. For the latter, the range of the a-particles ofradioactinium is somewhat longer than that of actinium-X, and itis to be expected, therefore, that its period of average life should beless instead of greater.It is possible that one of these two periods,either that of 28.1 days ascribed to radioactinium, or of 15 daysascribed to actinium-X, may be in error. Naturally this hasfocussed attention on these two products, and already a preliminaryannouncement has been made that radioactinium consists of twosuccessive products, the first having the period of average life of28.1 days hitherto ascribed to radioactinium, but emitting no, orbut little, radiation, and a subsequent product of period about19 hours, emitting a-, P-, and y-rays. The method of separation,however, is not stated.10The substances for which the ranges, but not the periods, areknown, are uranium-ZZ, radium-C’, and thorium-C’. For the twolatter the very short periods indicated by the long ranges of thea-rays are about 1 0 - 6 and 10-11 second respectively, and thereforeare quite beyond the present methods of direct determination. Thenew value for the range of the a-ray of uranium-2, 2.5 cm., fits thecurve for the series much better than the previous value.Thecalculated values for the periods of uranium-ZZ and ionium arerespectively 2 x 106 and 2 x lo5 years. I f these are correct, thereshould exist per gram of uranium about 1 milligram of uranium-ZZ,a chemically non-separable element of atomic weight 234.5, thepresence of which, however, will only affect the accuracy of theatomic-weight estimation of uranium in the third decimal place.The quantity of ionium in minerals should be about eighty times10 J.Chadwick and A. S . Russell, Nature, 1912, 90, 403RADIOACTIVITY. 295greater than the quantity of radium, or 24.6 grams per 1000 kilos.(25 grams per ton) of uranium. I n view of this the failure todetect the spectrum of ionium calls for comment Op. 321).Attempts have been made to find a theoretical basis for the a tpresent empirical relation between range and period.11 It is clear,however, that it will not be easy to do so. The ranges of the a-raysfrom any one product are all alike, whereas the actual periods oflife of the individual disintegrating atoms are widely different, themean only of all the values being constant. Yet the relation holdsbetween the range, which is truly a constant for each individualdisintegrating atom, and the period, which is only a statisticalconstant. Any explanation of the rule would undoubtedly throwlight on the nature of the process of disintegration itself.In viewof the increasing number of far-reaching deductions which dependon the rule, its present purely empirical character must not be lostsight of. The rule explains a t once why no a-particles of a veryshort range, as, for example, 1 cm., have been detected. I f it holdsgenerally, the life of the product emitting such a radiation wouldbe so long that the number of rays emitted would be far below thelimit of experimental detection, even by the most sensitive methods.The string electrometer has been successfully employed in placeof the usual quadrant type, in the work of counting, by the electri-cal method, the number of a-particles expelled from a radioactivesubstance.I n conjunction with a photographic arrangement forrecording the jerks of the electrometer mirror, great advances inthe accuracy and trustworthiness of the method have been made, asmany as 1000 a-particles per minute being photographically regis-tered. The method should enable the fundamental constant, QJ12the number of atoms disintegrating per gram of radium per second,to be determined to almost any required degree of accuracy. A t thesame time the apparatus has been improved in sensitiveness andsteadiness by the use of carefully purified helium instead of air asthe medium ionised, and it is expected that even the individualrecoil-atoms, the kinetic energy of which is about fifty times lessthan that of the a-partikle, will ultimately come within its rangeof detection.13The scattering of the a-particles by collision with the atoms ofmatter is of two types, which have been distinguished by the terms“ single scattering ” and “ compound scattering.” The first appliesto the result of the multitude of very small deflections suffered byl1 H.A. Wilson, Phil. Mag. 1912, [vi], 23, 981 : A., ii, 617 ; F. Butavand,Lc Radium, 1912, 9, 203 ; A , , ii, 723 ; R. Swinne, Physikal. Zeitsch., 1912, 13, 14 ;A , ii, 219.l2 Ann. Report, 1909, 234.l3 H. Geiger apd E. Rntherford, Phil. Mag., 1912, [vi], 24, 618 ; A., ii, 1021296 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.all the a-particles in successive encounters, and the second to theoccasional and comparatively rare deflections of the a-particlethrough large angles as the result of single collisions, on whichsuch important theoretical deductions have been based.It has beenshown that the scattering, experimentally observed, agrees well withwhat is to be theoretically expected.14 The theoretical variation ofrange due to the varying chances of collision of the individuala-particles with the gas molecules has also been discussed.15Some interesting results have been obtained in the microscopicexamination of the tracks of single a-particles from polonium in aphotographic film. Under high magnifications up to 1700 diametersthe straight tracks are resolved into straight rows of dark points,corresponding with the granules of the emulsion.It is stated thatnot all the granules in the direct tract of an a-particle are affected,but there seems to be some lack of clearness on this point, and theopposite could certainly be argued from other facts in the paper.If the a-particles before reaching the film are caused first to traversevarying distances of air, the length of the track in the film and thenumber of affected granules which constitute the track, bothdecrease in constant proportion to one another and linearly withthe distance of air traversed. The range of the a-particle in aircan thus be deduced by a new method from a photomicrographicstudy of the tracks.16A simple method of measuring and of demonstrating to anaudience the range of the a-particle by projection has been devised.It depends on the fact that an atmosphere laden with fumes ofammonium chloride and acted on by an electric field clears whenionised.I f one of the edges of a rectangular transparent chamberis coated with an a-ray-giving substance, an electric field beingmaintained between the opposite faces, and the chamber is filledwith fumes of ammonium chloride, these clear away rapidly overthe region penetrated by the a-rays, leaving a sharp dividing linebetween the cloud and the cloud-free space."/3- and y-Rays.With the complete elucidation of the physical nature of thea-rays, the attention of investigators has largely been transferred t othe &rays, which continue t o yield results of the highest theoreticalsignificance.The actual photographs which have been published1912, [vi], 23, 901.l4 H. Geiger, Proc. Boy. SOL, 1912, A, 86, 236; C. G . Darwin, Phil. May.l5 K. F. Herzfeld, Physiknt. Zeitsch., 1912, 13, 547.l6 W. Michl, Sitmtgssber. K. Akad. Wiss. Wien., 1912, 121, IIa, 1431.K Przibram, ibid., 221R ADlOACTIVITY. 297of the “magnetic spectra ” of line sources of &rays, resolved intoseparate beams by a magnetic field, are full of interest, and revealthe extraordinary complexity of the phenomenon.’* A t leasttwenty-nine sets of distinct &radiations have been recognised inthe photographs of the rays of radium-B and -C. The &rays ofuranium-X and radium-E and the swifter &rays of mesothorium-2and of thorium-C+-D have not yet been resolved, and theirmagnetic spectra consist of, or contain, broad bands, which mayor may not prove to be resolvable into distinct lines by morepowerful methods.I n the other cases the &radiation has beenresolved and for each product several distinct beanis have usuallybeen recognised. I n the products of the thorium series thirteendistinct beams and two bands exist. This is in striking contrastto the a-rays. Each a-ray-giving product yields only one type ofa-radiation, or, to put it even more explicitly, each disintegratingatom gives one a-particle, the initial velocity of which is, withinthe narrow limits of experimental detection, the same for all thedisintegrating atoms of the same kind.Both types of radiation,however, suffer a loss of velocity in passage through matter, whichfor the 8-rays is much greater for the slower than for the fasterrays.19 Passing the radiation through absorbing screens causes themagnetic spectrum to be spread out. Hence these kind of experi-ments can only be done with preparations of infinitesimal massas sources of B-rays, such as the active deposits, or the radiumemanation enclosed in tubes of minimal thickness, in which theabsorption of the radiation in the preparation itself is negligible.In view of the large number of distinct types of 8-rays thatexist, the number of B-particles emitted per atom disintegratingis of great interest. The researches so far made, by measuring thenegative charge transported by the &rays from preparations inwhich the number of atoms disintegrating per second can beevaluated, indicate that most probably only one &particle isexpelled from each disintegrating atom.20 The number, forexample, found both for radium-B and for radium-C was 1.1, forradium-E less than one, and for uranium-X, thorium-D andactinium-D, deduced from previous measurements, 1, 0.8, and 1.4respectively.I n any case it is clear that the number is far lessthan one for each of the numerous distinct types of &radiationrecognised. Using the Loreatz-Einstein theory for calculating theJ. Danysz and J. Gotz, ?:bid.,0. v. Baeyer, 0. Hahn, and L. Meitner, Physikal. Zeitsch., 1912,E. Rutherford, “RadioactiveJ. Danysz, Le Rncliwm, 1912, 9, 1 ; A., ii, 219.6 ; A , , ii, 220.13, 264 ; A ., ii, 409.Suhstances and their Rodiations,” Cainb. Univ. Press, 1912, p. 253.l9 Baeyer, Zoc. cit.0. v. Baeyer, ibid., 485.Fig. SOc.H. G. J. Moseley, Proc. Boy. Soc., 1912, A, $7, 230; A . , ii, 1014298 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY,kinetic energy of the 13-particle a t various speeds, the kineticenergies of the numerous successive groups of &radiation givenout by radium-C can be well expressed by a relation of the formpE,+pE,.21 The values of E, and E, are respectively 1.12 and0.356 ( x lO13e ergs), p and p being integers having values between0 and 9 and 0 and 2 respectively.For radium-B, the successive groups show a nearly constant dif-ference of energy X3? which is nearly one-half of E, (0.173 x 10%ergs).The theory has been advanced that the 8-particles from anyone disintegrating radio-element are, like the a-particles, initiallyall the same as regards speed, but that, unlike the a-particles, theysuffer, whilst still within the parent atom, successive decrements ofenergy which is utilised in the production of the y-rays.TheP-rays are known to be very easily deflected after expulsion bycollision with the molecules of matter, whereas the a-particlesplough their way for the most part straight through them. Onthis view the y-rays are regarded as analogous to the characteristicX-rays excited in the elements by catlode-rays of above a certaincritical velocity. A t the time the theory was proposed, theseradiations had not been experimentally produced in elements ofhigh atomic weight on account of the difficulty of artificiallygenerating cathode-rays of the necessary velocity.It is to beexpected, however, that such characteristic radiations would havea penetrating power closely in accord with that possessed by theknown y-rays, whilst the energy of the cathode-rays necessary toexcite them would accord with that possessed by the faster &raysof the radio-elements. The view adapts itself to the conception ofthe “Saturnian” atom, in which a very large positive charge isconcentrated as a single nuvleus a t the centre and is surroundedby rings of electrons external to it and rotating in planes. I ntraversing these external rings of electrons, the ejected &particlesuffers its successive decrements of energy, each of which correspondswith the production of a y-ray.Thus, in the case of radium-C,the emission of the &particles is accomplished by the productionof p y-rays of one kind and p y-rays of another kind. Thesesuggestions are, no doubt, to be considered as more or less tentative.If further investigations establish the fact that the decrements ofenergy of the &particle occur by definite quanta during its passagethrough the parent atom, and if though p + p y-rap are produced,not p+ p kinds of y-fays but two kinds only result, the complexityof the &rays and the apparent homogeneity of the y-rays wouldbe accounted for, although the structure of the parent atom pro-21 E.Rutherford, Phil. Mag., 1912, [vi], %, 453, 8931 A., ii, 1024 ; 1913, ii, 4RADIOACTIVITY. 299ducing this result would require further elucidation. At first sight,however, the complexity of the &rays and apparent simplicity ofthe y-rays, taken in conjunction with the existence of 8-rays ofhigh velocity unaccompanied by y-rays, to be considered later(p. 313), seems additional evidence in favour of regarding the twokinds of radiation as not necessarily causally connected.The changes which attend the character of the @-radiation duringpassage through absorbing materials still occupy investigators.The gradual reduction of velocity in aluminium without, a t least,the complete destruction of the homogeneous character of the beam,is well brought out in the photographs of the magnetic spectra ofthe &rays of the thorium active deposit after passage throughaluminium foil.= Screens of different metals retard the rays pro-portionally to the density of the metal. The same conclusion hasbeen reached in experiments on the absorption of &rays renderedhomogeneous by magnetic sorting.23 The radiation is a t firstreduced in velocity, with an increase of the specific ionising power.I n consequence, the ionisation is a t first increased by passage ofthe radiation through aluminium foil.Only after 0.04 cm. of foilhas been traversed is the ionisation reduced by the absorbingscreens. However, in passage through platinum a totally differentcourse is followed. The homogeneity of the beam is completelydestroyed, and the ionisation, even when absorbing screens ofaluminium are used, now decreases according to the usualexponential law.Probably too little account, in former experi-ments both with /3- and y-rays, has been taken of the specific actionof the absorbing material. Light metals, like aluminium, behavevery differently from heavy metals, like platinum and lead.The y-rays of radium are so superior in penetrating power toordinary X-rays that it is hardly possible to decide whether ornot the two radiations are of the same type. The y-rays obey the“density law” of absorption much more closely than X-rays, butof recent years the recognition of characteristic secondaryX-radiations has gone far toward explaining the peculiaritiesexhibited in the absorption of X-rays by different substances.Thegap between the two types of radiation may be considered to havebeen definitely bridged by a research in which, in the first place,it has been shown that when the very soft type of y-rays, such asthose produced by radium-E, are used for comparison, theabsorption phenomena encountered are in no way distinct fromthose exhibited by X-rays. Even more important is the proof thatthe y-rays of radium-E are capable of exciting the same charac.0. v. Bneyer, Zoc. eit. J. Danysz, Compt. rend., 1912, 154, 1502 ; A., ii, 617. * W. Wilson, Proc. Boy. Soc., 1912, A, 87, 310; A., ii, 1023300 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.teristic radiations as are excited by penetrating X-rays24 in theelements with atomic weights between silver and neodymium.Undoubtedly, in the course of time, the few anomalies encounteredin the absorption of the y-rays, in which the “density law” isdeparted from, will be rendered clearer by this conception of theexistence of radiations characteristic of the atoms of matter andindependent of the character of the exciting radiation, so long asthis exceeds a certain minimum of what, in acoustics or optics,would correspond with pitch or frequency, necessary t o stimulatethe atoms into radiation.Even a-rays, it would appear, possess the power of exciting thischaracteristic radiation.The preliminary announcement has beenmade that an excessively small but detectable amount of y-radiationis excited when the a-rays of radium4 impinge on matter.Theconclusion which follows, that all the radio-elements expellinga-rays must give a small amount of y-radiation also, is supportedby very recent experiments, in which it was found that a veryrich preparation of ionium emits some y-rays, but no detectable&radiation. Three types of y-rays were distinguished, with thevalues of p / D in aluminium 400, 8.2, and 0.15 (cm.)-1. These, i tis concluded, are probably characteristic radiations of ionium orthorium in three different series. The value for the middle series,8.2, corresponds roughly with that found for the (( L ” series excitedin thorium by X-rays.26Further results with 8- and y-rays are more conveniently dealtwith in the sections Multiple Disintegration and y-zay Methods ofMeasurement .&Rays.The low velocity electrons emitted by surfaces under the impactof a-rays are of general interest, because their formation has beenrecognised as the analogue in solids of the process which in gasesresults in ionisation.Certain conclusions seem gradually to have been placed on afirm basis.The &rays appear to be emitted with a velocity follow-ing Maxwell’s distribution law for gas molecules. No definiteupper limit can be assigned to this velocity for the same reasonas in the case of the velocity of the gas molecules, but for the6-rays from a polonium-coated metal plate, velocities correspondingwith a fall of potential of as much as from 33 to 43 volts have beenrecognised.With a-rays from another source bombarding a metalplate, the maximum recognisable velocity of the b-rays corresponds25 J. Cliadwick and A. 5. Eussell, Nature, 1912, 90, 463.J. A. Gray, Proc. €by. Soc., 1912, A, 8’7, 489RADIOACTIVITY, 301with 15 to 20 voIts, and the most probable velocity with 6 volts.From polished surfaces the electrons can escape under their owninitial energy of motion, but with rough surfaces, such as thosecoated with soot, an electric field is necessary to assist them toescape.26 Certain conclusions that the &rays, when first produced,possessed no definite velocity, their velocity being acquired sub-sequently to their formation, have now been withdrawn27It has been proved that X-rays behave similarly to a-rays inproducing b-rays, and a detailed investigation has established thatit is the secondary &rays, which X-rays generate on impact withsolids, which produce the &rays.If surfaces struck by X-rays arecovered with paper, very few P-rays are generated, and theb-radiation is also suppressed. The b-rays so generated are in everyrespect identical with those produced by u-rays. The feature ofperhaps the most interest appears to be that the speed of theb-rays is independent both of the nature and temperature of themetal in which it is generated and of the nature of the excitingradiations, and is to be ascribed to some mechanism common to allionisation. The incident and emergent 6-particles produced bya-rays in passing through a thin sheet of metal are also alike inall respects.28Evidence has, however, been obtained that the layer of adsorbedgas on the surface plays a certain part in the production ofb-radiation.When such surfaces are exposed to a-rays in a highvacuum, the &rays generated diminish notably in quantity withlapse of time. No such “fatigue” is noticed with surfaces fromwhich the film of gas has first been removed.29 The characteristicsof the b-rays are, however, in no way different in the two cases,and it is impossible, however completely the gas film is removed,completely to prevent the emission of 6-rays. Indeed, the numberof &rays is only reduced by about 30 per cent., a figure which iswithin the variation in the number of ions produced by the samea-rays in different gases.As is well known,30 the curves connectingionisation and range of a-rays have different forms in differentgases, when gases of widely different density are compared.Whereas the analogous 6-ray-range curves 31 have, unexpectedly,the same form when surfaces so different in density as those ofaluminium, copper, gold, lead and platinum are compared. Theemission of b-rays is thus not four square in all respects with the26 Fr. Hauser, Physikal. Zeitsch., 1912, 13, 936; A , , ii, 1026.27 N. Campbell, Phit. Mag., 1912, [vi], 23, 462; 24, 527 ; A, ii, 411, 1027.2d N. Campbell, ibid., 23, 46, 462; 24, 527, 783 ; A , , ii, 221, 411, 1027, 1121.29 Pound, ibid., 23, 813 ; 24, 401 ; A,, ii, 514, 886.30 Ann. Reporf, 1907, 312 ; 1909, 287.31 Ibid., 1911, 281302 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.production of ions.That all is not known is shown by the recentdiscovery of a new radiation itself completely absorbed by a singlealuminium foil, 0.64 x 10-4 cm. in thickness, corresponding withan air equivalent of only 0.58 cm., which causes the emission ofb-rays from surfaces on which it is incident. It has been suggestedthat part, if not all, of the b-rays and of the ionisation producedby ionising radiations is due to the intermediate production of thisnew radiation.32 Doubtless it accounts for many of the difficultiesexperienced in the investigation of the subject (see also RadioactiveRecoil).X-Ra y s.Space must be allowed for a brief reference to the excitingdevelopments now in progress as a result of the discovery thatthe X-rays are capable of showing up the actual space-lattice ofa crystal as theoretically conceived by crystallographers.33 A platewas cut from a crystal of zinc-blende, parallel to a cube face, andperpendicular to one of the principal cubic crystallographic axes.A narrow beam of X-rays was directed perpendicularly through iton to a photographic plate.On development, the plate revealed anumber of light spots on a dark background. A t the centre wasa circular spot 0.5 cm. diameter, surrounded symmetrically bysixteen smaller elliptically-shaped spots, arranged iu a square,each side consisting of four spots separated by 0.5 cm., the cornersof the square being free from spots.Other fainter systems of spotswith similar cubic symmetry, parallel or diagonal with the first,appear. One is inside, and the others are outside and a t con-siderable distances from the main square. These spots wereascribed to an interference diffraction photograph of the space-lattice of the crystal.The activity which followed the original discovery of the X-rayshas been revived.34 It has been shown that the X-rays, ‘which, ofcourse, do not obey the ordinary optical laws of reflection, arestrongly and regularly reflected in accordance with these laws whenthey impinge on the cleavage planes of a crystal a t nearly grazingincidence. Crystals of rock salt and mica sheets have been used,and it is probable that this type of reflection, from the internalcrystal planes, accounts fully for the diffraction patterns of thespace-lattice obtained.It is almost certainly not surface reflection,a H. A. Bumstead and A. G. McGougan, Phil. Mag., 1912, [vi], 24s 462 ; A., ii,53 M. Laue, W. Friedrich, and P. Knipping, Sitzzcngsber. K. Ahad. Miinchat,Compare A. E. H. Tutton, Naluw, 1912, 90, 30;.34 W. H. Bragg, Xature, 1912, 90, 219, 360; W. L. Bragg, ibid., 410 ; C. G.1026.1912, 303, 363.Barkla and G. H. Martyn, ibid., 435RADIOACTIVITY. 303although a mica strip only 0.1 mm. thick reflects as strongly as athick crystal. By bending the mica into an arc, the X-rays canbe brought to a line focus. These discoveries cannot fail to exerta profound influence on the search for a theory, which shallembrace equally all the properties of light, X-rays and y-rays, andreconcile the conflicting aspects of the wave and corpuscularhypotheses.The Cloud Method of Re~izdering the Tracks of Rays Visible.This interesting new method, described last year,Jb has quitefulfilled the expectations that it would prove to be an exceedinglypowerful means of settling some of the most difficult questions asto the connexion between ionising radiations and the ionisationthey produce. A fine selection of photographs, comprising thetracks of u- and /%rays, and of the secondary 8- and cathode-raysproduced by beams of y- and X-rays, has been published, togetherwith full details of the vastly improved methods employed in theirproduction.36 In the first place, the whole clear conception of thenature of the a-rays, with their almost straight paths, their sharplydefined range, the increase of the ionisation as the ray penetratesthe gas and is slowed down, the columnar ionisation, the (‘single”and ‘( compound ” scattering, which previous researches have dis-closed, is shown to be accurate to the minutest particular in thesevisible records.Minor new points of great interest also appear.At the “ single scattering ’’ which usually characterises the end ofthe path, whenever the angle of deviation is large, a backwardshort extension of the new path, or (‘spur,” appears. This isprobably due to the atom which has been struck and which causesthe sudden abrupt deflection, itself acquiring sufficient energy toionise for a short distance.A t the beginning of the path of ana-ray, when it is possible to see the beginning, as in the a-rayexpelled from an atom of emanation, for example, an enlargedhead appears. This probably marks the ionisation due to therecoil-atom of radium-A , which, also, has been otherwise demon-strated by recent experiments.Likewise, with regard to the &rays, the general view as to theirionisation and scattering has been confirmed. It is possible to seeon the photographs the siugle pairs of drops corresponding withindividual pairs of ions, and even to count the number producedper cm. of path, with results in fair agreement with existing data.In addition to the regularly spaced pairs, clusters of 20 t o 30 dropsalso appear, which indicate that some of the atoms struck them-35 Ann.&port, 1911, 275.36 C. T. R. Wilson, Proc. Roy. SOC., 1D12, A, 81, 277304 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.selves liberate cathode-rays possessing energy sufficient to ionise forshort distances. The scattering is almost wholly of the cumulativeor compound type. Gradual and considerable curvature, ratherthan abrupt changes of angle, characterise the tracks of &rays soloiig as their velocity is high.The X-ray photographs show no trace whatever of directionisation apart from the secondary cathode-ray liberated from theatoms of the gas, and thus fully bear out the contention thatX-rays only ionise indirectly. These cathode-rays are of muchlower velocity than the P-rays of radium, and show both kindsof deflection, the gradual or cumulative deviation being muchthe more important factor in scattering.But abrupt deflections,often through 90° or more, also occur, especially a t the end of thepath. The cathode-rays start apparently in all directions, and norelation between these directions and that of the primary X-raybeam has yet been made out. The appearance of these photographsis that of a mass of interlaced fine threads, no one of which isstraight, all being both curved and bent characteristically. Onclose examination, and with enlarged pictures, the individual dropsof the clouds and the dotted appearance of the tracks are clearlyseen.Ionisa tion.That ionisation is not fundamentally atomic, but depends tosome extent on chemical combination, is recognised as one of thevery few points of contact between the new subjects dealt with inthis report and the older molecular sciences.Thus sulphur dioxideis a denser gas than hydrogen sulphide and absorbs X-radiation ofall degrees of penetrating power to a greater extent in consequence,yet the relative ionisation in layers of equal thickness is 1-24 timesgreater in the latter than in the former. I n a mixture of oxygenand hydrogen sulphide the ionisation is 1.17 times greater than ina mixture of hydrogen and sulphur dioxide of the same com-position.37 The variation in the number of ions produced indifferent gases by a-rays has long been recognised. A recent com-parison of the ionisation-range curve in mercury vapour a t 330°and 451 mm., with that of air a t the same temperature and a t apressure adjusted to make the ranges equal, showed that theionisation produced in mercury is nearly 40 per cent. greater thanin air.% Yet in these cases the molecular ionisation has beenrecognised as mainly, at least, an additive property of the atoms3’ C.G. Barkls and L. Simons, Phil. Mag., 1912, [vi], 23, 317 ; A., ii, 222.38 T. 5. Taylor, ibid., 24, 296 ; A., ii, 888RADIOACTIVITY. 305of the molecule,39 and not a constitutive property as in the caseof X-ray ionisation just cited.The number of pairs of ions produced by the single a-particle ofpolonium in air has been estimated to be 164,000, the mean energyabsorbed in producing one pair of ions, obtained by dividing theenergy of the a-particle by this number, being 5.3 x 10-11 erg.40The extraordinary results obtained from the study of positiverays in vacuum tubes, in which a host of novel types of molecularaggregations have been recognised,41 belong properly to the subjectof Gas Analysis, and cannot be dealt with in detail.Some interest-ing conclusions as to the true nature of ionisation are, however,put forward. The atoms in a molecule are probably not oppositelycharged. Thus, in a molecule of hydrogen or oxygen, the twoatoms are not charged respectively with positive and negativeelectricity. Each atom is, more probably, electrically neutral,containing equal numbers of opposite atomic charges of electricity.The chemical affinity is conditioned by the disposition of thesecharges with reference to one another.We may infer from thisthat a single bond holding two similar atoms in a diatomic moleculeis not well represented by the conventional symbol, but it neces-sarily double, such as might graphically be represented by 5o:nesuch symbol as:This is of interest because a molecule so constituted would possessone degree of freedom less than that with the conventional singlebond.An important paper dealing with the mathematical theory ofionisation by moving electrified particles can only be mentionedhere.42y-Ray Methods of Measwement,I n dealing with the radio-elements, the proportion of the elementin different preparations and therefore the relative atomic weightof the element can be determined exactly by methods for the firsttime essentially different from those so familiar in the cese ofordinary elements. I n the case of radium, f o r example, the ahinicweight, in terms of those of the other elements involved, couldbe obtained with fair accuracy by y-ray measurements without asingle weighing on a number of different compounds, such as thechloride, bromide, carbonate, and so on.y-Ray measurements3@ Ann. a p o r t , 1907, 314. 40 Phit. Mag., 1912, [vi], 23, 670 ; A , , ii, 412.41 Sir J. J. Thonison, ibid., 24, 209 ; A., ii, 885.43 Ibid., 23, 449 ; A., ii, 410.REP.-VOL. IX. 306 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.would invariably be employed, by preference, because theabsorption of y-rays in the substance itself can readily bearranged to be negligible.Thus, y-ray measurements andthe accuracy that can be attained through them are of con-siderable interest. A t Paris in the comparison of the radiumstandards two main methods were employed. In one,43 the radiumtube was placed on the centre of a large horizontal circular plateof lead, 1 cm. thick, which formed the upper plate of a condenser,the distance between the plates being only a few centimetres. Theupper plate was connected to one pole of a battery of 800 volts,the other pole of the battery being earthed, and the lower plateis connected with the one pair of quadrants of an electrometer,the other pair being earthed. A t a given moment an earthconnexion of the lower condenser plate is broken, and the ionisationcurrent between the condenser plates is exactly neutralised by theCurie quartz electric balance, in which a weight of several kilo-grams, held in the hand of the operator, is gradually applied tothe pan of the instrument so as to keep the needle of the electro-meter as near zero as possible.With practice this apparentlydifficult operation can be accomplished with marvellous precision.As soon as the full weight is applied, the needle swings away fromthe zero off the scale, and the time is taken. The weight employeddivided by the time gives the relative measure of the strength ofthe y-rays. The absolute measure of the ionisation current can,if required, be readily deduced from the constant of the electricbalance, and does not involve the capacity of the system.In the second method44 a form of optical bench was employed,a t one end of which a small, shallow, cylindrical ionisation chamberof lead was mounted with the axis parallel to the bench, the radiumpreparations being supported in the line of the axis on a lightaluminium stand sliding on the bench.The ionisation currentdue to the y-rays of radium was balanced against the ionisationcurrent due to a surface of uranium oxide, kept constant duringthe comparison. Balance was obtained by sliding the radiumpreparation along the bench. In comparing two radium pre-parations X and P, if A and B denote the balance distancesrespectively, measured to the inside surface of the nearer end ofthe ionisation cylinder, and I is the length of the cylinder, theratio of the two preparations X/ P =B(B + Z)/A(A + I ) .Thisratio is subject to a very small correction for the different absorptionof the y-rays in the air a t the two distances.The two methods gave concordant results with the various radium‘3 A. Debierne, Le Badium, 1912, 9, 169.4 E. Rutherford and J. Chadwick, Proc. Physical Soc., 1912, 24, 1-41 ; A., ii, 520RADIOACTIVITY. 305standards employed, and the error of measurement was estimatedas certainly less than 3 parts in 1000. As an example, the numbersfound for the 31.13 and 10.11 milligram Vienna standards bycomparison with the Paris standard were, in single determinations,31.24 and 10.13. I n absence of other substances giving y-rays,the quantity of radium in a sealed tube can now be estimatedwithout opening the tube with a t least an accuracy of 0.3 per cent.The second method has been used to measure the absorptioncoefficient of y-rays of radium in gases and light substances.45 Thevalues of p / D , where p is the absorption coefficient and D is thedensity, for y-rays after traversing 0.3 cm.of lead are 0.048,0.051, and 0.047 for air, carbon dioxide, and hydrogen respectively,and for the first two for y-rays after traversing 1 cm. of lead,0.046 and 0.047. For air at Oo and 760 mm., .the value givenfor p is 0-0000624.46 The ‘‘ average path” of the y-ray in air isthus about 150 metres. For a great variety of substances of widelydifferent densities, from that of wood to that of lead, p / n variesfrom 0.0401 t o 0.0599, and has a minimum value for substancesof intermediate density.It is greater both for lighter and heaviersubstances, the difference being the more pronounced the lesspenetrating the y-rays employed, as found previously.I n another research the absorption coefficient of the &rays ofradium in air has been directly measured.47 For radium pre-parations sealed in glass test-tubes and for short ranges, 0.6 t o1.6 metres, of air, p is 0*0033(~m.)-~, and for longer ranges, 2 to5 metres, 0.0045. The value obtained for the y-rays was, however,less than half that given above. The interesting and possiblyimportant observation was made that, for the &rays from a baresurface of radium active deposit, the value of p, is about 0.012.This value does not alter with time, when the proportion ofradium-C t o radium-B is increasing in the active deposit.Thisindicates that radium4 as well as radium-B emits a considerableproportion of very easily absorbed &rays.Solutions have been published of some important mathematicalproblems in the absorption of y-rays in materials of various form.48Helium and Neon.In an investigation of the mineral waters of the King’s Well,Bath, the interesting observation was made that the proportions45 J. Chxdwick, Le R d i ~ n ~ , 1912, 9, 200 ; A., ii, 718.46 Compare also V. I?. Hess, Xitzungsber. K. Aknd. Wiss. Wien, 1911, 120, IIn,1205, where the valne found for p was 4-47 x 10-5(cm.)-1.47 A.S. Eve, Trans., Roy. SOC. Canada, 1911, 5, iii, 59 ; A , , ii, 717.H. Thirring, Physikal. Zeitsch., 1912. 13, 266. E. v. Schweidler, ibid., 453.L. V. King, Phil. Mag., 1912, [vil, 23, 242.x 308 ANNUAL REPORTS ON THE PROGRESS OF CREMISTRY.of argon, neon, and helium therein differ notably from theatmospheric proportions, and that hydrogen and oxygen are absent.There is 0.78 times as much argon, 188.1 times as much neon, and72.8 times as much helium as in the same volume of atmosphericair. The water contained 1.73 x 10-9 curie, and the evolved gas33-65 x 10-9 curie of emanation, t u t only 0*14xlO-g gram ofradium was present in the water, per litre.49The absence of oxygen and hydrogen, such as might be expectedfrom the radioactive decomposition of the water, and the greatpreponderance of neon is taken as indirect evidence that neon isproduced by the action of the emanation on the water.Neon insuch a quantity, however, is an exceptional constituent of mineralspring gases, whereas radioactivity of the degree of the Bathwater and higher is comparatively common.Neon, together with helium, was found to be present in thegases produced from a solution of thorium nitrate, which had beentreated with several quantities of radium emanation. The coldcharcoal method of separating the helium a.nd neon was employed,whereas a chemical reagent, such as lithium or calcium, which doesnot absorb the atmospheric argon, if present, is to be preferred incases where the origin of a small quantity of neon is a t issue.Theunabsorbed gases had a volume of 0.485 cu. mm., and the spectrumshow both helium and neon brilliantly. It was judged from theappearance of the spectrum that from one-third to one-fourth, thatis, about 0’177 cu. mrn., consisted of neon. The gases absorbed inthe cold charcoal were treated for argon, but hardly any wasrecovered. The conclusion was arrived a t that the air-leakage wasinsignificant, and that the neon must have been produced by theaction of the emanation on the solution.The conclusion, however, is hardly borne out by the evidence.For it was found that 0.5 C.C. of air, which contains 0.006 cu. mm.(not 0.06, as stated), gave a spectrum ‘(comparable in brilliancywith, but not so brilliant as, that of the neon separated from thegases of the thorium solution,” which from spectroscopicexamination had been adjudged to contain 0-177 cu.mm. of neon.I n a preliminary communication 60 it is stated that neon has beenobtained by bombarding fluorite, prepared by precipitation, wash-ing, and heating to dull redness, with cathode-rays in a tube towhich oxygen is admitted as required to maintain the pressure a tthe value best suited for the production of cathode-rays. Thesurface turns purple, and silicon fluoride, oxygen, and carbon49 Sir William Rainsay, Chem. News, 1912, 105, 133 ; A . , ii, 417 ; T., 1912, 101,1367. J. I. 0. Masson and Sir W. Ramsay, ibid., 1370.Sir William Ramsay, Nature, 1912, 89, 502RADIOACTIVITY. 309monoxide are expelIed.The gases evolved during the first fewdays were rejected, and tbe fifth quantity pumped off was examined.After absorption of the condensable gases by cold charcoal, theresidue was pure neon, withollt a trace of helium.Researches on the proportion of the rare gases in French naturalsprings show that, with the exception of helium, the other membersof the family occur in proportions of the game order as in theatmosphere. The conclusion, with regard to neon, is, however,provisional only, owing to the difficulty of accurately estimatingthis constituent. The springs of the CGte d’Or contain 10 percent. of helium in the nitrogen and rare gas fraction, whilst theSource Carnot a t Santenay is equally rich, and is estimated t oyield 17,845 litres of helium annually.51 The question arises as tothe origin of this helium, for the quantity furnished by the SourceCarnot requires for its steady annual production some 70,000 kilos.(about 70 tons) of pure radium, equivalent to some hundreds ofmillions of kilos.of rich uranium minerals. The same questionarises with regard to the natural gas in certain parts of America,several of which are stated to contain over 1 per cent. of helium,62which must be evolved in enormous quantities.Tiiermo-radioactivity .With the large quantity of radium chloride of definite puritypossessed by the Radium Institute of Vienna, determinations havebeen made of the heat generated and other constants. Underconditions in which the whole of the a- and &radiation and anestimated fraction of 18 per cent.of the y-radiation were absorbedin the calorimeter, the heat generated per hour per gram of radiumchloride was 100.66 calories, or per gram of radium, 132.26 calories.These figures refer to radium in equilibrium with its first fourshort-lived producfs.53 The energy of the rest of the y-rays wouldincrease the figure by an amount, still somewhat indefinite, butprobably from 3 to 5 calories per gram of element. The valuesagree so well with the figures calculated from the known data asregards the kinetic energy and number of the a- and @-particlesand of the recoiling atoms, that the very important conclusion isarrived a t that no appreciable part of the energy of the radiationescapes transformation into heat.No appreciable part, that is tosay, is absorbed in doing any other kind of work than that involvedin raising the temperature of the surrounding matter.A further determination of the heat generated by radium, freed51 C. Moureu and A. Lepape, Conapt. rend., 1912, 155, 197 ; A., ji, 833.5L Ann. Beport, 1907, 323.53 St. Meyer and V. F. Hess, Monatsh., 1912, 33, 683 ; A . ii, 716310 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.from all its disintegration products,54 gave the value 25.2 caloriesper gram of radium per hour, and, as in this case there is nopenetrating radiation, the figure is definite. This leaves 107.1calories for the heat generated by the emanation, radium-A, -B,and -C, under condifions such that the whole of the a- and Praysand 18 per cent.of the y-rays are absorbed. What must be con-sidered as a truly remarkable result is that this figure, 25.2 calories,agrees within 1 per cent. with the value calculated from Joule'sequivalent, and the number, mass, and kinetic energy of thea-particles expelled from radium, taking into account the energyof recoil, and assuming that the whole of the kinetic energy sufferstransformation into heat.An independent analysis of the heat in calories generated perhour per gram of radium in equilibrium with its first four productsis given in the following table as the result of an exhaustiveexperimental investigation with the radium emanation and itsproducts, interpreted in the light of the other data applying tothese products.66a-Rays.&Rays. ?-Rays. Total.Radium ................. 25 -1 - 25 -1Emanation.. ............. 28.6 - 28.6Radium- A ............ 30.5 - 30.5Radium-B + -C ......... 39'4 4.7 6'4 50 *5Total ............ 123 '6 4.7 6-4 134'7- --The heat from a tube containing a given quantity of radiumemanation in equilibrium with its products was first measured, theemanation was removed as quickly as possible, and from the changewith lapse of time in the amount of heat generated, it was deducedthat of the total heat 29, 31, and 40 per cent. respectively camefrom the emanation, radium-A and radium-B+-C. From thevelocities of the various a-rags, the proportion should have been28.8, 30.9, and 40.3 per cent., but taking into account the factthat probably about 4 per cent.of the total energy is due to0- and y-rays which come from radium-B and -C only, the heatgenerated by radium$ appears somewhat less than theory indicates.Whether this small difference is significant remains to be seen.On the assumption that the heat effects are proportional to theionisation produced by the rays, an attempt was made to estimatethe proportion of the total heat due to p- and y-rays, and fromthese experiments the figures in the table are deduced. They differfrom preceding estimates 66 in the increased proportion ascribed tothe &rays. This is because the measurements were done in theF*l V. F. Hess, Sitzungsber. K. Akad. Wiss. W i e n , 1912, 121, IIa, 1419.65 E. Rutherford and H. Robinson, ibid., 1491.56 Ann.&port, 1911, 283RADIOACTIVITY. 311special glass tubes capable of allowing even the a-rays to escape,whereas in ordinary tubes, even the thinnest, a large proportionof the softer @-radiation is absorbed.It is estimated that the polonium in equilibriuni with 1 gramof radium will generate a further 25.86 calories per hour, theamount produced after one year's accumulation being about1 calorie per hour. The calculated saturation current (in termsof E.S.U. x 106) from a gram of radium, free from products, in theform of an infinitely thin film so that there is no absorption, is1-28 (when half of the radiation only escapes to produce ionisation),whereas that experimentally found was 1.22. The total saturationcurrent calculated for 1 curie of emanation in vessels of infinitedimensions is 2.75, and for the emanation together with its productsof rapid change, 6-10, these figures agreeing closely with the variousexperimental determinations, extrapolated to correspond withvessels of infinite dimensions from the formula of Duane andLaborde.57 It is clear that the standards of radium employed inthe past by different investigators do not differ very widely fromthat adopted as the International standard.58A determination of the heat generated by a specimen of orangite,estimated by a radioactive method to contain 36 per cent. ofthorium, gave in two series 25.4 and 19.4 ( x 1 0 - 5 calorie per gramper hour), which is about ten times that found for thorium oxideby other observers, and is greatly in excess of that indicated byradioactive theory.Possibly some chemical change in the materialmay have caused the result, but until the repetition of the experi-ments now in progress is complete, discussion may be reserved.59Multiple Disintegration.Thorium-C or -C, and -C,.-The most definite case of multipledisintegration a t present appears to be that of thorium-C. Adopt-ing the view that each mode of a dual disintegration obeys the lawof a single radioactive change as though the other mode did notoccur, at rates indicated by the radioactive constants A, and A,respectively, the parent substance divides itself so that the fractionh1/hl + A, disintegrates according t o the first mode and the fraction]\z/hl + according to the second.The radiation decays exponen-tially according to the equation I = 1,e - ( A i + W , independently ofwhether the rays result in both modes or in one only of the twomodes of disintegration.6057 A m . Rcport, 1910, 256.58 H. Itlache and St. Meyer, Physikal. Zeilvch., 1912, 13, 320 ; A . , ii, 580.Mcyer and V. F. Hess, Monalsh., 1912, 33, 588 ; A., ii, 716.59 H. H. Poole, PhiZ. Mug., 1912, [vil, 23, 183.60 F. Soddy, ibid., 1909, [vi], 18, 739 ; A., 1909, ii, 952.St312 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.If A, and A, are quantities of the same order, as appears to bethe case with thorium-C, the process is much easier to follow experi-mentally than if A, and h2 are of widely different value, as appearsto be the case with radium-C. From t.he general manner in whichthe radioactive constants vary widely from member to memberand from series to series over an extreme range at present knownof 1 : 103O, it is to be expected that the cases in which A, and h2happen to be similar will be few in comparison with those in whichthey are widely different.There is little doubt that many of thscases of disintegration a t present regarded as simple may proveon closer examination to be multiple, with A, and A, widely different.A single parent element may thus give rise to many end-productsvarying enormously in relative amount, the ratio A,/A, fixing alsothe ratio of the relative amounts of the two end-products in a dualdisintegration.For every 100 a-particles given by the thorium emanation, 100are given by thorium--4, whilst 35, of range 4-55 cm., and 65 ofrange 8.16 cm.are given by thorium-C, indicating that thorium-Cundergoes a dual disintegration with hl / %= 35 / 65 .61 In the caseof the a-particles of range 8-16 cm., the period estimated by theGeiger-Nuttall relation is 10-11 second.Thorium-C disintegrates according to an exponential law withperiod of average life 1 / ~ = 7 9 minutes, so that, on the viewtaken, l / A l + A,= 79 minutes, and the separate values of h, andA, would be 225.7 and 121.5 minutes. It is these separate periodsand not the apparent period which would fix the range of anya-particles emitted in the changes. The first period agrees wellenough with the range (4.55 cm.), but it is clear that the longerrange a-particle (8.16 cm.) must come from a change with periodof the order of 10-11 second, and this change must follow the121.6 minutes period change of thorium-C.In the first place, very numerous attempts to detect any lack ofhomogeneity in thorium-C were unsuccessful.Under all circum-stances, the 35/65 ratio between the two sets of a-particles holdsgood, and this is maintained unaltered as the radiation decays.Thorium-C, which had been separated by recoil from the activedeposit, by volatilisation a t various temperatures and by immersingplates of various electrochemical potentials in the solution of theactive deposit, always gives this same ratio. Similarly, when veryshort exposures to the emanation were given, the active depositbehaved normally, the a-radiation rising to a maximum in twohucdred and twenty-five minutes, and maintaining the ratiobetween the two sets of a-particles unchanged.All this is evidence61 T. Bnrratt, Proc. F'h.gslsicn2 sbc., 1912, 24, 112 ; A . , ii, 408RADIOACTIVITY. 313that thorium-C is a homogeneous type of matter, the atoms ofwhich can disintegrate in two distinct ways.62 Just aa, accordingto present ideas, there is no inherent difference recognisable betweenatoms of the same radio-element which are about to disintegrateimmediately and those which are destined to survive possibly severaltimes the period of average life, the selection following the laws ofprobability, so in thorium-C there seems to be no recognisable differ-ence between those atoms which will expel low-range u-rays (C,)and those which will expel high-range u-rays (Cz), the laws ofprobability controlling the number following each mode as thoughit alone took place.A most interesting point has, however, transpired.Per atom ofthorium-C disintegrating one and not two a-particles are expelled,and therefore one &particle also is expelled.63 I f thorium-D were,%or I urn 3eries.as previously supposed, the product solely responsible for the &rays,and if i t gave, as in other cases, one &particle per atom disintegrat-ing, it would follow that thorium-D must result from both modesof the disintegration of thorium-C. This means that the series afterbranching must join together again, which is in the highest degreeimprobable.A closer examination has brought to light the follow-ing new facts. All the y-rays, but only a portion, the relativelysmaller portion, of the &rays, come from thorium-D. The greaterpart of the B-rays, some 65 per cent., come from thoriumC, and areunaccompanied by y-rays. The &rays expelled by the C memberare considerably swifter and more penetrating than those expelled62 E. Marsden and C. G . Darwin, Proc. Roy. Soc., 1912, A, 87, 17 ; A., ii, 823.63 Ann. Report, 1911, 282314 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTHY.by the D-member. Hence it is reasonably concluded that thedisintegration runs as shown in the diagram on p. 313.The three circles lettered C refer to the same substance, disin-tegrating dually, both modes of disintegration proceeding simul-taneously, not successively.The hypothetical, shortclived parent ofthe 8.16 cm. range a-ray has been designated C" for want of abetter term. The periods refer to the periods of average life. Thefacts as regards the 8- and y-rays are well established by entirelyindependent investigation^.^^ The mixed &rays of thorium8 + -Dare half absorbed in 0.41, of pure -D in 0.32, and of pure -C (calcu-lated) in 0.48 mm. aluminium. The &ray curve of pure thorium-Crises from an initial value 0.72 to a maximum, reckoned as unity, inten minutes, whereas the y-rays rise from an initial value of only0.06 to a maximuin in fourteen to sixteen minutes.Thus the dual disintegration of thorium47 occurs either as anu-ray or a &ray disintegration.I f the a-ray is first expelled, theproduct gives P- and y-rays, whereas if the &ray is first expelledit is unaccompanied by y-rays, and the product disintegrah withimmeasurable rapidity, giving a-rays. Here we have evidence forthe first time of penetrating l3-rays unaccompanied by y-rays.Evidence, however, has been published of a partial separation bychemical methods of thorium-C into two numbers responsible forthe two kinds of a-radiati~n.~S I f this is correct, it completely upsetsthe preceding explanation. Since the evidence is indirect, and allthe details of the methods of separation are not fully stated, it may,theref ore, be left over, witshout prejudice, for subsequent considera-tion, when fuller evidence is forthcomiDg.The explanation hereadopted fits all the known facts provided thorium-C is homogeneous.Should it prove separable into distinct products, -C, and e2, thenthe dual disintegration must occur earlier, for example, a tthorium-B. But in this case thorium-C, and -C2 must have the sameperiod. Otherwise, the a-radiation would not decay exponentially.The relative number of a-particles emitted by the successive pro-ducts shows that a dual disintegration does occur.In the whole thorium series six u-particles are expelled per atomof thorium; the calculated atomic weight of the end-product istherefore 232.4 - 23.94 = 208.46, and the nearest known atomicweight is that of bismuth. Two end-products, however, with thesame atomic weight in the ratio 35 : 65 must 'result, if both branchesof the series end as the diagram shows.Badium-C, and -C2.-There is not much new experimental evidence64 E.Ifarbden and C. G, Darwin, Zoc. c l t . 0. Hahn and L. Meither, PhgsikEnZ.65 L. Meitner, PhysikaZ. Zeitsch., 1912, 13, 6 2 3 ; A., ii, 723. Compare alboZ~zlsch., 1912, 13, 390 ; A,, ii, 514.von Hevesy, PhiE. Mag., 1912, [vi], 23, 628; A . , ii, 414RADIOACTIVITY. 315in the case of radium-C, but an explanation analogous to that ofthe thorium dual disintegration may be put forward.66 I n this casethe name radium-C, has, unfortunately, been applied to a distinctproduct, of period of average life 1.9 minutes giving &rays, but nota-rays, separable in very small quantity only by recoil fromradium-C, that is, from the complex product of period 28.1 minutes,giving a-, /3-, and y-rays.At first the minuteness of the quantityrecoiled was attributed to the substance being produced by a &raychange. But in other &ray changes, for example, that of radium-E,no trace of the product has been observed to recoil. Adopting theview that the relative minuteness of quantity of radium-C, is realand not merely apparent, it appears that radium-C undergoes adual disintegration with the ratio hl/h2 about 3/10,000. That is,some 3 atoms out of every 10,000 of radium4 may be supposed to-0 28.1 rninRadrum Seriesdisintegrate, with expulsion of an a-radiation, of a range which canbe calculated to be about 3.9 cm., but which has not yet beendetected, producing radium-C2, which then expels &rays, andpossibly y-rays also.The remaining 99-97 per cent. of the atomsexpel j3- and y-rays, and produce a product of estimated period10-6 second, which on disintegration gives the long range a-raysand produces radium-D. In the diagram the hypothetical shortrlived product has been termed " C'," but it is clear that the wholenomenclature needs to be recast if it should be proved that thethorium and radium series are analogous. These diagrams mayprove to be premature. At least, they are interesting.Evidence in favour of the real relative minuteness of quantityK. Fajans, Physikal. Zeitsch., 1912, 13, 699; A . , ii, S24316 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.of radium-C, is obtained from the study of the time-b-ray curves ofpure radium-B, in which the proportion of 8-rays contributed byradium-C2 was too small to be detected.67Incidentally, these experiments confirmed the conclusion ofSchmidt, which had been called in question, that radium-B emitssome penetrating P-radiation.About 1.5 per cent. of the ionisationis due to &rays [p = 13(cm.)-lAl], identical with those of radium-C,the remainder being due to the @)-rays [ p = 91 (cm.) -lAl] commonlyascribed to radium-B. No evidence of still softer rays (p.890)found by Schmidt was obtained.Following this up, it was found that radium-B also emits y-rays,and 12.7 per cent. of the ionisation produced by the y-rays ofradium-B and -C in equilibrium were found to be due to radium-B.The absorption-coefficient in lead between 0.3 and 0.6 cm. was4 (cm.)-l, and between 0.97 and 1-72 cm., 6 (crn.)-l.These y-raysare thus of very feeble penetrating power compared with those ofradium-C, and through 2.3 cm. of lead no y-rays due to radium-Bcould be detected.Radioactive Recoil.A recoiling atom possesses some fifty times the mass and one-fiftieth of the velocity and energy of the a-particle from which itrecoils, but, like the u-particle, it appears to be capable of ionisingthe molecules of the atmosphere in which it is produced.68 Probablyrecoil atoms constitute part of the new radiation referred to atthe end of the section &Rays. I n the case of radium-C, the energyof the a-particles is very great, and the recoiling atoms of radium-Dpossess a “range” one-five hundredth of that of the a-rays ofradium-C, or about 0.14 mm.of air at atmospheric pressure. Atdistances within this from a surface of radium-C, the recoiling atomsof radium-D produce five times as much ionisation as is producedby the a-rays. The fraction of the total ionisation contributed bythe recoiling atoms is, however, less than 1 per cent. So far ascan yet be seen, the ionising power of the recoil atoms diminishesas the distance traversed in the gas increases, which is the converseof the law holding for a - r a y ~ . ~ ~Similar experiments with polonium (Ra-F), from which the stillunidentified and hypothetical end-product of the series must recoil,were unsuccessful until a preparation of the requisite degree ofpurity was prepared to give a layer thin enough (less than 10pp) toallow the recoil-atoms to escape.Then the ionisation due to the67 K. Fajans and V;. Makower, Phil. Mag., 1912, Ivi], 23, 292 ; A., ii, 220.6a Ann. aport, 1910, 273.69 L. Wertenstein, Le Radium, 1912, 9, 6 ; A., ii, 222RADIOACTIVITY. 317recoiled atums became manifest, and the range was estimated as7 cm. in air a t 1 mm. pressure, or rather less than 0.1 mm. atatmospheric pressure. The ratio of the ranges of the recoil-atomsof Ra-C and Ra-F are nearly the same as the ratio of the rangesof their a-particles.70 This is the first experimental evidence of theexistence of an ultimate product of a disintegration series, asrequired by the theory.The absorption of recoil-atoms in passing through a gas has beeninvestigated in the case of radium-C.The range of the recoilingatoms of radium-B is 10.5 cm. in air a t 1 mm. pressure. Hydrogenat 6 mm. pressure is equivalent in every respect to air a t 1 mm.I n air a t 1 mm. the diminution of the number of atoms in acanalised stream of recoil-atoms is unimportant for the first 5 cm.,and then increases rapidly t o the end of the range. In hydrogen at6 mm. the diminution is much more abrupt, the lighter atoms ofhydrogen probably scattering much less than in the case of air. I neach case the ionising power of the recoil-atom diminishes alongits path, but the diminution is much more rapid in air than inhydrogen.The amount of recoil product (actinium-D w'as the productstudied) collected on a negatively charged surface is diminishedby intense ionisation of the gas.The recoiling atoms behave inthis respect like positive ions. Increase in the concentration ofthe ions in the atmosphere increases the chance of their recombina-tion with ions of opposite sign before they arrive a t the electrode.71ZnfZuence of Temperature on Radioactive Changes.The Existence of Chemical Compounds of Short-livedRadio-elemen ts.A complete explanation has been obtained of the small changesin intensity of the less penetrating types of radiation during theheating and cooling of tubes containing the radium emanation andactive deposit.72The effects are not a t all due t o real changes in the radioactivity,but to alteration of the position of the radiating materials withrespect to the measuring instrument by volatilisation and condensa-tion, whereby the mean distance of the source of rays from theinstrument and the fraction of the radiation absorbed by the wallsof the containing vessel is altered.73 Thus, if a spherical bulb isemployed it can readily be understood that the absorption of the70 B.Bianu and L. Wertenstein, Compt. rand., 1912, 155, 4 i 5 ; A., ii, 887.T1 A. F. Kovarik, Phil. Hag., 1912, [vi], 24, 722 ; A., ii, 1121.72 Ann. Repod, 1910, 273.73 A. S. Russell, Proc. Roy. Soc., 1912, A, 86, 240 ; A., ii, 416318 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.radiation in the walls must be less, if the active matter is all in thestate of gas, than if it is all deposited uniformly on the walls.I nthe latter case a greater proportion of the rays reaching an externalinstrument are tangent to the surface of the bulb, and thereforehave paths passing through a maximum thickness of the wall.Surprising variations occur, however, in the degree of volatilityof the various numbers of the (‘ active deposit ” groups in differentexperiments, which were traced to the nature of the atmospherein which volatilisation took place.74In a vacuum, actinium-B, deposited on a surface of platinum orquartz, begins to volatilise a t 600°, and volatilisation is neaflycomplete at 900°. I f the deposit is first exposed to various gaseousreagents, such a6 chlorine or hydrogen iodide, its volatility isincreased.Hydrogen chloride, whilst not affecting the volatility,increases the amount of actinium-B which can subsequently be dis-solved by water, the actinium-C being hardly at all soluble underthese conditions. I n experiments in which a long mica strip wasinserted in a tube hotter at one end than the other, a wire coatedwith actinium-B being placed a t the hot end, condensation of thevolatilised actinium-B occurred in air mostly a t places heated above1000°, but in hydrogen the maximum amount was deposited on thesurface at a temperature becween 615O and 675O.Similarly with the radium active deposit, in hydrogen volatilisa-tion of all three products, radium-A, -B, and -C, is complete a tabout 650°, but in an atmosphere containing oxygen no volatilis&tion of radium4 occurs below 1200°, and neither of the threeproducts is volatile below 700O.In hydrogen volatilisation ofradium-C commences even a t 360O. For this substance there isthus a difference of 800° in the volatilisation temperature inhydrogen and oxygen respectively.75These remarkable differences point strongly to the existence ofdifferent chemical compounds of these short-lived radioelements ;thus in hydrogen the substance may exist in elementary or metallicform, which is relatively volatile compared with the oxide formedin presence of oxygen. The whole direction in which the subjectis moving shows that it is by no means so impossible as appears atfirst sight to investigate the chemical nature of these ephemeralsubstances, although they only exist in quantities below themillionth of a milligram.74 H.Schrader, PJriZ. Mag., 1912, [vi], 24, 125 ; A . , ii, 722.i5 A. S. Russell, ibid., 134 ; A., ii, 723RADIOACTIVITY. 319Electrochemistry of the Radio-elements.The foregoing remark applies aptly to the electrochemistry ofthe radio-elements, for in this subject a standing reproach has beenthat the theory often deals with concentrations entirely beyond therange of verification, as is instanced, for example, by the electro-chemical determination of the solubility of almost insoluble sub-stances. Applied t o the radieelements of short-lived period inwhich a detectable radioactivity is associated with a billionth of amilligram by weight, the electrochemical theory leads to the resultthat such substances should be deposited from solution in detectableamounts by differences of potential far below the decompositionvoltage.In a neutral solution some detectable quantity of a shortrlived radio-element should be deposited on any metal, and thistheoretidal deduction has been verified for all metals, including thenoble metals gold and sirver, in the case of fifteen separate short-lived raclio-elements.76Using concentrations of the order of 10-16 g. p0r c.c., the v a r ktion of the ratio in which the -B and -C members of the seriesdeposited was investigated, as the electrochemical potential of themetal was varied from the cas8 of silver in silver nitrate solution,where the metal is positive t o the solution, to that of zinc in zincsulphate, which is negative.I n the first case the C-member isdeposited almost pure, but as the potential of the metal decreasesand changes sign the B-member is deposited in increasing quantityuntil it is in excess of the equilibrium ratio. The B- and C-membersof the three series exhibit identical electrochemical behaviour, andare deposited in equilibrium ratio when the metal is 0.63 voltnegative to the solution.Not the least interesting development was a radioactive methodof determining single electrode potentials, for which electrochemicaltheory cannot be directly tested, from the ratio of the -B and -Cmembers deposited, and the known variation of this ratio with theP.D.actinium-B and -C furnishes the quickest, and thorium-B and -Cthe most accurate indicator for this purpose, radium-B and -C beingunsuitable. The ratio is deduced from the a-ray decay curve of thedeposit; thus for a copper electrode immersed in pure water forone minute the potential difference between metal and solution was-0*7V., a-nd for an immersion of 0.2 second, -2V. With longerimmersion the normal electrochemical potential of copper isapproached. These are beautiful illustrations of how deductionsfrom the theory of solution pressure in cases where the ionicconcentration of the metal approaches zero may be tested.7776 G. von Hevesy, Phil. .Wag., 1912, [vi], 23, 628 ; A., ii, 414.77 G. von Hevesy, Phpsikal. Zeikc?h., 1912, 13, 715320 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Von Lerch's general method of separating the -C member fromthe -B member in a pure state by deposition on nickel dependspartly on the abnormal negative potential of the nickel electrode,which is connected with its passivity, and on the use of an acidsolution in which the -B member deposited is preferentially dis-solved.In the three series radium-C is the easiest and thorium-Cis the most difficult to separate from the -B member in the purestate, whereas from the molecular equilibrium ratio in the threeseries, actinium-C should be t h s most difficult of the three. Thisanomaly would be explained if thorium-C really consisted of twoseparate substances, C, and C,, of which only the one, C,, givingthe.shorter range a-rays and present in less proportion, is analogousto the other C-members. The point is, however, too crucial t o besettled by such indirect evidence.To nium .Homogeneity o f Chemical Elements.The results of experiments that have been in progress over thepast s v e n years on the growth of radium from large quantities ofuranium in solution purified in various ways have recently beencollected and discussed.78 A minute growth, which, expressed inunits of 10-12 gram of radium per kilogram of uranium per year,varied from 12.4 for the first t o 2.0 in the last quantities purified,was observed in all the uranium preparations, but, so far as canbe seen, it is proportional to the time and is due to minute tracesof ionium originally present.No indication of an increased rateof growth of-radium, such as must occur if ionium in appreciablequantity was forming during the time of observation, has yet beenobserved. It may be calculated that, if ionium is the only long-lived intermediate substance in the series between uranium andradium, its period of average life must be a t least 100,000 years,which is forty times longer than that of radium, and the quantityin uranium minerals must be a t least 12'3 grams per 1000 kilos. ofuranium. The total quantity of ionium and thorium together inJoachimsthal pitchblende is such that the maximum estimate forShe period of ionium can hardly b0 greater than a million years,and has recently been estimated by this method on certain assump-tions to be not more than 250,000 to 300,000 ~ears.7~ These esti-mates are supported by the failure to detect the growth of ioniumfrom uranium-X, the product of which remains experimentallyunknown.80 They all involve the assumption that no unknown7* F.Soddy, Royal Institution Lecture, Mnrch 15th, 1912 ; Xnlurc, 1912, 89, 203.70 St. Meyer and V. F. Hess, Sitzrcngsber. K. Aknd. Wiss. Wiea, 3912, 121, 11. ;d., ii, 603.Ann. Report, 1909, 262RADIOACTIVITY. 321intermediate substances of long life intervene, On the other hand,the period of 200,000 years, estimated from the range of the ioniuma-particle (p. 294), does not involve this assumption, and is ingood agreement with the other estimates.The question, whether unknown long-lived intermediate membersof the series still remain to be discovered, has, however, been raisedin an acute form by the failure t o detect, in the arc spectra of thestrongest ionium-thorium preparations, a single new line due toionium.81 Such preparations, from the intensity of their a-radia-tions, must contain more than 10 per cent.of ionium if the periodis 100,000 years, and, it is reasonably certain, should show at leastsome of the lines of the ionium spectrum. The strongest line ofthe radium spectrum can be detected in a barium preparationcontaining only a few parts of radium in a million. Not a singleunknown line has, however, been observed. I n one investigationin addition to the thorium spectrum, only five of the stronger linesof scandium were observed.In the other investigation, cerium,scaiidium, yttrium, and some of the rare earths were detected. Itmay be mentioned that the proportion of ionium and thorium in apreparation of the oxides is not affected by heating in the electricarc.The com-plexity of “uranium” and the view that it consists of two non-separable elements, uranium-Z and uranium-ZZ, differing in atomicweight by four units, raises the question whether uranium-X, theproduct of “uranium,” is the product of uranium-Z or ofuranium-ZZ. There is really no experimental evidence whatevercapable, at present, of deciding between the two alternativeschemes :Ur-I -> Ur-11 --P+ Ur-X -!+ 10 -> etc.Ur-I-> Ur-X -!+ Ur-II -> 10 > etc.An interesting line of thought may now be pursued.Hitherto, the first has always been assumed, but the second ismuch the more probable.For the first scheme is an exception tothe rule that the expulsion of an a-particle reduces the valencyby two units, the step being always from the family of evenvalency into the next, the family of odd valency being alwaysmissed.82 The second scheme not only conforms to this rule, butexhibits also the well-marked alternation, shown in other cases, inwhich the next change, in which an a-particle is not expelled, causess1 F. Exner and E. Haschek, Sitzungsber. K. Akad. Wiss. Wien, 1912, 121,IIa, 1075 ; A. 5. Russell and R. Rossi, Proc. Roy. Soc., 1912, A, 87, 478.52 Soddy, “ Chemistry of the Radio-Elements,” p. 30.REP.-VOL. IX. 322 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.the atom to revert t o its original groupfollowing comparison of the uranium and thorium series :Th -'?+ MsTh -5 RaTh -> Th-X -> Eiii -> etc.IV.I I. I v. 0.This is shown by theUr-I -> Ur-K -!+ Ur-ZI -> 10 -'$ Rn -$ Em -> etc.The figures refer to the families in the Periodic Tables to whichthe elements belong. All the members in both series designated bythe same numeral (0, 11, IV, and VI) are chemically identical, sofar as is known.83Obviously, if uranium-ZZ, and not ionium, is the product ofuranium-X, the experiments above referred to with the lattersubstance do not furnish a minimum estimate of the period ofionium, as hitherto supposed, but of that of uranium-ZZ. Thereasoning, however, as regards the absence of growth of radiumfrom uranium remains unmodified, for the two uraniums neces-sarily always exist together in equilibrium proportion.The simplest, if somewhat heterodox, view t o take is that ioniumis a long-period element, and that its spectrum, as well as its wholechemical behaviour, is identical with that of thorium.It is clearthat the conception of chemical elements as necessarily homogeneousis undermined, and that different elements with different atomicweights are chemically identical, not as an exception, but as aconsequence of the way in which the known disintegration serieshave been shown t o run their course. In non-radioactive matterthis heterogeneity cannot be distinguished. Not so in the case ofmatter actually in the process of evolution.VI.IV. VI. I v. XI. 0.Chemical Action of the Rays of Radium.In an investigation of the ozonisation of oxygen by a-rays thenumber of ozone molecules formed was found to be of the sameorder as the number of pairs of ions produced. The ratio betweenthese two quantities varied capriciously, and was in no case greaterthan 1 : 2, frequently being less. The results were consistent withthe view that the ozone results in a reaction between the oxygenions and molecules, but that unknown causes produced a loss of theozone formed, which varied according to the conditions.@The statement that radium emanation exerts a decomposing effect83 Compare A. Fleck, Paper read before Section B, British Association Meeting,Durtdee, 1911 ; Chem.News, 1912, 106, 128. The full results, obtained in theauthor's laboratory, cre in course of publication. Some of them bear on thesubjects discussed in this section.S. C. Lind, Monatsh., 1912, 33, 295; Le Radium, 1912, 9, 104 ; A, ii, 513RADIOACTIVITY, 323on sodium urate, converting it into easily soluble substances, hasbeen submitted to an exhaustive examination, and disproved.85Many of the reactions accelerated or brought about by light 01ultraviolet light have been examined under the influence of thepenetrating rays, of radium. The effects produced by the latterare often very small by comparison. No increase in the rate ofesterification of benzoic acid by alcoholic solution of hydrogenchloride, no effect on p-benzoquinone in alcoholic or etherealsolution, and no action on normal aqueous solutions of oxalic acidat 25O, could be observed, although very powerful radium prepari+tions and long exposures were employed.The formation of theacid from o-nitrobenzaldehyde in alcoholic or benzene solutionswas accelerated by radium rays, although to an extent between10,000 and 20,000 times smaller than is produced by ultravioletlight from a quartz mercury lamp a t 8 cm. distance. As with ultra-violet light, the rays of radium effect a marked increase in therate of inversion of unsterilised neutral sucrose solutions, but theyact oppositely to ultraviolet light in favouring the growth of mouldin sugar solutions.86The liberation of iodine from solution of alkali iodides by thepenetrating rays of radium is notably increased by presence of freeacid, as though the alkali produced had a retarding action, and,like most photochemical actions, is not much influenced by increas-ing the temperature.Potassium iodide is far more sensitive to thisreaction than sodium iodide.87 Similarly, a t least in acid solution,bromine is liberated from solutions of potassium bromide, but theaction is less marked than with the iodide. Acidified potassiumchloride solutions are unaffected.The velocity of decomposition of hydrogen peroxide is greatestin acid and least in alkaline solutions, and no difference wasobserved when paraffined glass vessels were employed. Ferricsulphate, especially in the presence of sucrose, is reduced by pene-trating radium rays as it is by light.88Biocheni ical Effects of Badioactiuity.Three important'communications deal with the differences broughtabout by radioactive agencies in the development of plants, thesprouting of buds of different woods, and the growth of seeds.In95 E. v. Knaffl-Lenz and W. Wiechowski, Zeitsch. physiol. L'hm., 1912, 77, 303 ;A . , ii, 522.bfi A. Kailan, Sitzirngsber. K. Aknd. 1Viss. Wien, 1912, 121, IIa, 1329 ; A . ,1913, ii, 8.87 A. Kailan, Monatsh., 1912, 33, 71; A , , ii, 522.~3 Ibid., 1329 ; A., 1913, ii, 7.Y 324 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.the first,89 watering with radioactive water from Joachimsthal, of300 to 2000 Mach units strength, some seven varieties of plants(Triticum vulgar'c, Hordeum disticum, Vicia faba, Pisium satiuum,Lup*nus angustifolias, Trifolium pratense, and Pisium aruense)produced after a week strikingly favourable effects on the growth.The roots and stalks increased several times in length, and theweight of dried matter in the stalks was from three to five timesgreater than in the case of plants watered with inactive waterof the same chemical composition.On the other hand, variousbacteria (13. mycoides, fluorescens lipuefaciens, pyocyaneum, andfilefaciens) were harmfully affected by the active water, whereasA zohacter chroococcum, the nitrogen assimilating bacterium, wasnot so much affected.I n the second paper,g0 it was found that the winter buds ofvarious woods such as Syringa uulga?-is, when exposed for one ortwo days to the rays of strong radium preparations, or, better,under a bell-jar in an atmosphere containing a few millicuries ofemanation, sprouted when subsequently cultivated in the hot-housein the light, whereas other buds not treated did not sprout at allor only much later.Too little exposure to the rays is withouteffect, and too much exposure is harmful or fatal to the plant. A tother periods of the year no result is produced, and the effect isquite distinct from that produced on the growing plant.Lastly,Ql the influence of the radium &rays of different penetratring power on the development of small seeds, under 0.7 mm.diameter, has been studied. It was found that the growth washindered by irradiation with 6-rays from 8 milligrams of radiuma t a distance of 1 cm., independently of the chemical constitutionof the seed (starch and fat content), which varied widely, and wasgreater for the less penetrating than for the more penetrating rays,when rays of equal ionising power were compared. The retardationof the growth is not noticeable for short exposures (five hours).Aft.er the fifth hour for the first day, the retardation increasesrapidly with the exposure, and then increases less rapidly for longerexposures up to three or four days.It is possible that very shortexposures have a favourable influence on Sinapis nigm and Panicurngermanicum.Various.It is convenient to collect together a number of observations ofEmanations.-The actinium emanation is dissolved by variousHans Molisch, Sitzungsber.K. Akad. Wiss. Wien, 1912, 121, I, 121.special rather than general interest.@ J. Stoklasa, Conqt. rend., 1912, 155, 1096.91 E. D. Congdon, if&., 1911, 120, IIn, 1327RADIOACTIVITY. 325liquids and by charcoal very much as in the case of the otheremanations, but is rather more soluble. The coefficient of solu-bility in water is 2, that of the thorium emanation being 1, andof the radium emanation, 0.33.92 There is no difficulty in detectingthe presence of actinium in a uranium mineral by means of itsemanation, provided small containing vessels and rapid currents ofair are employed.93 The diffusion of the actinium emanation hasbeen further studied, and compared with that of the thoriumemanation.All that can be said is that the molecular weightsof these two emanations must be nearly eq~a1.9~,4ctive Deposits.-Two researches have been made on the dis-tribution of the active deposit of radium under various conditionsbetween the positive and negative electrodes.95 The proportion ofthe active deposit finding its way to the cathode varies much inthe same way as the ratio of the ionisation current in the gas tothe saturation current under the same conditions. To account forthe deposition of the active products on the anode, it is suggestedthat the positive charge on the atoms of radium-A a t formationis neutralised, and in the case of some 2 per cent. reversed bycombination with negative ions in the gas.Negative Results.-No radioactivity is produced in such metalsas bismuth, antimony, and iron by the application of a highfrequency alternating magnetic field,Q6 nor in various chemicalreactions, such as the decomposition of hydroferrocyanic acid byheat, the reduction of potassium dichromate by heating with oxalicacid and of potassium permanganate by heating alone or withreducing agents.97 The radioactivity of rubidium and potassiumsalts is unaffected by exposure to light. No a-rays could be detectedfrom these compounds, and the thermal effect of their radioactivityis too small to be observed.98As regards the existence of radio-uranium, no evidence exceptnegative evidence is forthcoming.99Uranium-P also must be classed with the substances the existenceof which is in need of confirmation.92 G.von Hevesy, Physikal. Zeitsch., 1911, 12, 1214 ; A., ii, 117.9y Idem., ibid., 1213 ; A,, ii, 116.g4 Miss M. S. Leslie, Phi2. Mag., 1912, [vi], 24, 637; A,, ii, 1032 ; J. C.95 E. M. Wellisch and H. L. Bronson, ibid., 23, 714 ; A., ii, 521 ; G. Eckmann,96 J. H. Vincent and A. Bursill, Proc. Physical Soe., 1912, 24, 71 ; A., ii, 417.y7 W. A. D. Rudge, Proc. Camb. Phil. SOC., 1912, 18, 465; A . , ii, 519.98 E. H. Biichner, Proc. R. Akad. Wetemch. Amsterdam, 1912, 15, 2 2 ; A., ii,it9 H. Sirk, Monalsh., 1912, 33, 289; A . , ii, 519.McLennan, ibid., 370 ; A . , ii, 589.Jahrb. Xadioaktiv. Elektronik, 1912, 9, 157 ; A., ii, 620.724326 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Radioactivity of the Earth and Atmosphere.The fusion method of estimating radium in rocks1 has beenused in a systematic revision, characterised by the feature that,instead of samples of individual rocks, mixed samples composed ofa large number of specimens of different rocks of the same classwere dealt with.Only undifferentiated materials are included.These results establish a substantial difference between the radiumcontent of igneous and sedimentary rocks, which had been ques-tioned by another investigator. I n terms of units of 10-12 gramof radium per gram of rock, the means for igneous rocks, dividedinto three classes, are: (1) acid, 3.01; (2) intermediate, 2'57;(3) basic, 1-28, whilst for sedimentary rocks the mean is 1.4,excluding the calcareous sediments for which the mean is O*8.2The mean of the previous results obtained by the solution method,excluding Joly's results, which are higher, are distinctly lowerthan these new determinations, being, for the three classes ofigneous rocks, 2-17, 1-28, and 0.58 respectively.If 2 is taken as a mean quantity of radium for all the rocks ofthe crust, and 2 x 10-5 gram of thorium per gram of rock isincluded, the heat generated is 25 x 10-14 calorie per gram persecond, and a depth of 17 kilometres would supply all the heatreaching the surface of the earth from the interior.The increaseof temperature at the base of this 17 km. layer would be only2 4 6 O under equilibrium conditions. I f the internal heat is to beentirely attributed to radioactive changes and a state of thermalequilibrium is postulated, it is necessary to suppose that the radio-active layers are less rich than the average surface materials andextend to a greater depth than 17 km.For as Joly has deducedin a general manner, a given quantity of the radio-elements willgenerate a t the base of the layer a higher equilibrium temperaturethe more they are diluted with inactive material, and the greaterthe depth of the layer in consequence.So far as investigations have proceeded, there does not appearto be any systematic variation of the radium content of rocks withthe depth from which they are taken, and rocks from a greatvariety of localities exhibit results closely in accord with the generalmean.3A number of investigatione deal with the emanation content ofAnn.Report, 1911, 293 ; compare E. Ebler, Zeitsch. Elektrochem., 1912, 18,J. Joly, Phil. Mag., 1912, [vi], 24, 694 ; A., ii, 1032; A. L. Fletcher, ibid.,E. H. Buchner, Proc. K. Akad. Welensch. Amsterdzm, 1912, 14, 1063 ; A . , ii,532 ; A . , ii, 723.23, 279 ; A., ii, 224.525RADIOACTIVITY. 327air drawn from the underground soil a t various depths, and theescape of emanation into the atmosphere. It is clear that thequantity of emanation in the atmosphere is continuously main-tained by the amount escaping from the soil. The air drawnfrom even a metre or less below the surface of the ground is severalthousand times richer in radium emanation than the atmosphere,a mean value in one series being 2 x 10-lo curie per litre. Thesupply of emanation is, as is to be expected, not exhausted by thecontinuous withdrawal of the air. Even so, it has been estimatedthat only about 1/70th of the emanation generated by the radiumin the soil escapes into the underground air, and this also is whatis to be expected from the known small proportion of emanationwhich ordinarily escapes from solid substances containing radium.For the purposes of comparison, it may be stated that the 5000litres of gas escaping daily from the Bath springs are estimatedto contain 1.7 x 10-4 curie of emanation. Volume for volume, thisis about 200 times richer in emanation than the underground air,for example, in the neighbourhood of Cambridge and Dublin.The escape of emanation into the atmosphere, and the consequentimpoverishment of the underground air in emanation, is facilitatedby a strong wind and hindered by frost and rain, but the fluc-tuations of the barometer, apart from accompanying stornis andrain, is without direct influence.4In a voyage from Valparaiso to the East Indies, the emanationcontent of the sea water was found to change to an extraordinarycontent with the locality, increasing with the specific gravity andtemperature of the water. The active deposit in the atmospherewas probably not derived from the emanation in the ocean, butfrom emanation carried from the land by winds.5The question of the origin of the earth's penetrating radiation,whether it comes from the earth or the atmosphere, is attractingattention. Most investigators believe that in the free air thepenetrating rays come from the earth, although very near thesurface the effect of a superficial active deposit from the atmospherecan be detected.6 It is noteworthy, however, that in two recentballoon voyages, one a t night and the other during the day, inwhich a specially constructed thick-walled Wulf electrometer wascarried, a t heights f o r the most part about 200 metres from theearth, although the maximum height was about 1000 metres, nodefinite diminution of the effect due to penetrating radiation wasL. B. Smjth, Phil. Mag., 1912, [vi], 24, 638 : A., ii, 1031 ; J. Satterly,Proe. Camb. Phil. Soc., 1911, 16, 336, 356, 514; A., ii, ll?, 118, 522.W. Knoche, Physikal. Zeitsch., 1912, 13, 112, 152 ; A., ii, 223.A. Gockel, Jahrb. Radioaktiv. Elektronik, 1912, 9, 1 ; A, ii, 416 ; W. Strong,Terrestial Magnelism, 1912, 17, 49328 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.observed.7 A t 200 metres from the earth's surface, reckoned as auniform source of y-rays of radium of infinite extent, the absorptionof the air should reduce the intensity of radiation as much as alayer of 1.83 cm. of mercury, and this should, of course, be verynoticeable.8 Probably, however, the instrument was not sufficientlygood to use for the purpose. As is recognised by everyone whohas attempted to reduce the natural leak of an electroscope to itsabsolute minimum, the effect of the radioactivity of the walls isthe chief contributor to the natural leak unless very special means,which it is not wise to leave to the instrument maker, are adopted.This seems to have been the case in the present instance. Theinfluence of the varying amount of ballast carried by the balloonhas also been alluded to.QThe spectrum of Nova Genzinorum (1912), which shows a generalresemblance to that of the solar chromosphere, contains lines whichhave been ascribed t o radium and its emanation. This has ledto the spectrum of the solar chromosphere during the eclipses of1898, 1900, 1901, and 1905 being re-examined, with the resultthat agreement has been found between some of the lines andthose of the radium spectrum. The evidence, however, is con-flicting, and the fact that radium is the source of these linesappears extremely difficult to bclieve.10FREDERICK SODDP.7 V. F. Hess, Sitxicngsbcr K. Akad. Wis8. Wkn, 1911, 120, IIn, 1575.8 An exact expression for the absorption is given by L. V. King, Phil. Mag.,9 L. TT. King, Zoc. cit.lo Astronont. Xachr., 4582, 4589, 4600.1912, [vi], 23, 211