August 28, 1913] 



NATURE 



659 



LETTERS TO THE EDITOR. 

 [The Editor does not hold himself responsible for 

 opinions expressed by his correspondents. Neither 

 can lie undertake to return, or to correspond with 

 the writers of, rejected manuscripts intended for 

 this or any other part of Nature. No notice is 

 taken of anonymous communications.] 

 Radium-D and the Final Product of the Radium 



Disintegration Series. 

 A recent discussion in the columns of Nature has 

 raised the question of the existence of elements 

 chemically and spectroscopically identical but differing 

 from one another by a few units of atomic weight. 

 Should the far-reaching generalisations of Soddy and 

 of Fajans prove correct, the chemical elements will 

 have to be regarded from an entirely different point 

 of view, and the explanation of the periodic nature of 

 their properties will have to be sought for in the 

 recurrent character of the processes of radio-active 

 disintegration. At present, however, some of the 

 evidence on which those views are based is at least 

 open to question. With the exception of radium, 

 niton, and the parent elements — uranium and thorium 

 — none of the thirty-four radio-active bodies have been 

 isolated, and none of their compounds have been 

 obtained in a state approaching purity. Their proper- 

 ties have only been inferred from their behaviour 

 when mixed with very large amounts, comparatively 

 speaking, of known elements, such as lead, thorium, 

 and tellurium, and when it has been found impossible 

 to alter the proportions of these minute traces of 

 radio-active matter by the usual analytical processes, 

 chemical identity has been inferred. Even admitting 

 the extreme delicacy of radio-active methods of 

 analysis, it might be questioned whether much can 

 be deduced from the inseparability of such minute 

 traces when present along with large amounts of 

 closely related elements. More satisfactory evidence 

 would be furnished if the chemical properties of an 

 equilibrium mixture from a pure radio-active parent 

 could be investigated ; and since during the last few 

 months some direct observations have been made on 

 th. nature of radium-D and its products it may be of 

 interest to give a preliminary account of the investiga- 

 tion. 



The source of the radium-D and its products was 

 highly purified niton (radium emanation) which four 

 years ago had been compressed into fine capillary glass 

 tubes, liquefied, and used for the determination of 

 the vapour pressures of the gas by Sir William Ramsay 

 and myself. Each tube originally contained somewhat 

 less than 0-2 of a curie of emanation, which, when 

 liquefied by pressure, filled a volume of approximately 

 1/5000 cub. mm. at the sealed end of the tube. Three 

 tubes of the precious material were kindlv placed 

 at my disposal by Sir William Ramsay. The 

 emanation was allowed to decay under pressure, and 

 subsequent microscopic examination showed that the 

 liquid had transformed itself into a dark-coloured 

 deposit of submetallic lustre, resembling somewhat a 

 dried colloidal metal, and also into a colourless gas, 

 viz. helium. The total mass of the solid deposit was 

 hence of the order of 1/1000 milligram, and since it 

 had decayed for four years it contained, in addition 

 to the equilibrium quantities of radium-E and 

 radium-F, about 15 per cent, of radium-G, the end 

 point of the series, as well as traces of the branch 

 series derived from radium-C„. The tube was only 

 weakly radio-active, but when laid on a photographic 

 plate the impression produced bv the /3-radiation of 

 radium-E corresponded exactlv with the visible dis- 

 tribution of the deposit in the tube. Bv a suitable 

 procedure the mercurv sealing the open end of the 



NO. 2287, VOL. qi] 



tube was removed as completely as possible, and after 

 pumping off the helium pure chlorine was admitted. 

 At ordinary temperatures the submetallic deposit re- 

 mained unattacked, but on gently warming it was 

 seen to change completely into a pure white, apparently 

 homogeneous, crystalline chloride. On heating 111 

 vacuo to 220 C, the chloride did not volatilise per- 

 ceptibly, and the photographic impression it produced 

 coincided with that obtained from the submetallic 

 deposit. Hence the chlorides of radium-D, -E, and -G 

 are not appreciably volatile under these conditions. 



On introducing water into the tube it was seen 

 that the crystalline deposit dissolved without decom- 

 position, but was only slightly soluble. Even after 

 warming in presence of about ten times its own 

 volume of water only a small proportion of the total 

 solid went into solution. 



The chief object, however, of this experiment was 

 to determine whether radium-G, the final product of 

 the series, was similar in chemical character to lead, 

 and also to find out if radium-D approximated closely 

 in behaviour to its longer-lived descendant. That 

 radium-G and lead are identical is supported by much 

 indirect evidence, though no direct proof has been 

 advanced. Now lead is an element which can be 

 detected in very minute quantities by the delicate and 

 characteristic microscopic test of Behrens, viz. by the 

 formation of a characteristically crystalline triple nitrite 

 with the nitrites of copper and potassium. The test is 

 so delicate that 1/100,000 milligram of lead can be 

 detected with certainty if the proper conditions are 

 observed, and, moreover, by determining the number 

 and size of the crystals in a drop of known volume 

 the amount present can be approximately estimated. 

 It was proposed, therefore, to apply the test to a 

 known fraction of the radio-active matter in the tube 

 and to see whether the amount of triple nitrite formed 

 corresponded with radium-G alone or with radium-G + 

 radium-D. The application of the test was compli- 

 cated, however, by the discovery that the glass of_ the 

 capillary, and, in fact, soda-glass in general, contained 

 about 0-03 per cent, of lead, and that a detectable frac- 

 tion of tliis lead could be dissolved out of the glass 

 with nitric acid. By avoiding the use of strong acids and 

 simply extracting the glass with water no perceptible 

 amount of this element could be found in the concen- 

 trated extract. To be quite certain that no lead from 

 the glass could find its way into solution by this 

 procedure, the very stringent test was made of extract- 

 ing 5 grams of finely powdered glass for some hours 

 with" water and testing the residue after evaporation to 

 dryness. The glass used had at some previous time 

 been exposed to the action of radium emanation, and 

 was of a deep purple colour. No lead was found in 

 the extract, though analysis proved its presence in the 

 glass. 



Other sources of error lay in the possible presence 

 of lead in the mercury sealing the tube, and in the 

 reagents used for the test. No lead could be detected 

 in the mercury and the reagents were carefully purified 

 beforehand. In order to carry out the test the capillary 

 tube containing the radio-active matter was cut into 

 two portions, a longer one containing most of the 

 solution, and a shorter portion in which the un- 

 dissolved crystals remained. The solution in the 

 longer portion was allowed to evaporate on a silica 

 microscopic slide, and one portion of the drop ob- 

 tained was tested for lead by the triple nitrite test , 

 and another tested with potassium chromate. In 

 both cases the presence of lead was indubitably proved, 

 but the quantity present was small. The short end 

 containing the crystals was then fractured and the 

 fragments extracted with water to which a drop of 

 acetic acid had been added. The evanorated extract 



