Theodore W. Richards (1868-1928) and Max E. Lembert

The Atomic Weight of Lead of Radioactive Origin

excerpted from Journal of the American Chemical Society 36, 1329-1344 (1914)

Received May 16, 1914.
It has been pointed out by many of the workers upon radioactivity, especially by Boltwood, Ramsay, Rutherford, and Fajans, that the most conclusive test concerning the recent theory of the degeneration of radioactive elements is to be found in the determination of the atomic weights. If each α-transformation involves the loss of an atom of helium and nothing else which is weighable, the atomic weight of the product should be just 3.99 less than that of the original substance, because 3.99 is the atomic weight of helium evolved during the α-transformation. Thus, if radium has an atomic weight of 225.97,[1] its emanation ("niton") should have an atomic weight of 221.98, radium D (which is supposed to involve three more α-transformations) should be 210.01; and radium G (yet another α-transformation) should be 206.02.

Still more recently, a further theory, which has been independently proposed by Fajans, and by Soddy,[2] indicates that some of the places in the periodic table, corresponding to high atomic weights, should perhaps each include several elements, different in atomic weight but very similar in other properties. Thus, in the place which we usually assign to lead, we should expect to find a mixture of ordinary lead, radium B, D, and G, and perhaps, also at least one other radioactive product from thorium and one from actinium. These different substances, according to the hypothesis, should have identical spectra and be inseparable by chemical means, but, coming from different sources, they should have different atomic weights. The theory supposes that each α-transformation involves a loss of valence of two, and each β-transformation a gain of valence of one. The β-transformation involves no change of weight. Thus radium D (supposed to have an atomic weight of 210) after two β- and one α-transformations returns again as radium G to the same place in the periodic system with an atomic weight of only 206. This place is that assigned to lead (which some suppose to be primarily radium G), the only one of the radium series possessed of a long life and not highly radioactive.

The problem is one capable of a decisive gravimetric test; specimens of lead, consisting of different atomic weights. On the generous suggestion of Dr. Fajans this matter was taken up in the autumn of 1913 at Harvard.[3] In order to glean as much knowledge as was within reach, we have endeavored to obtain as many different samples of radioactive lead as possible and to determine the atomic weights of the possibly composite element. by precisely comparable methods, so as to discover if any variation might exist in the chemical equivalents of the different products.

It is a pleasure, at the outset, to express our deep gratitude to many workers in radioactivity who have furnished us with material. Without this general coöperation, it would not have been possible for us to accomplish anything in so short a time, and we cannot express too highly our appreciation.

In brief, the method of analysis was essentially similar to that used so successfully by Baxter and Wilson in their work upon the atomic weight of ordinary lead.[4] The chloride was in each case prepared in a state of great purity by recrystallization in quartz and platinum vessels, after extensive preliminary treatment to eliminate foreign substances. This chloride was carefully dried in a dessicator and heated to fusion in a stream of hydrochloric acid gas and nitrogen, in the quartz tube of the well-known bottling apparatus which has served in so many similar cases.[5] The lead chloride was then dissolved in much water, and the chlorine precipitated by silver nitrate. Both the weight of the silver required and the weight of the precipitate were determined in the usual Harvard fashion.

As a further check upon the work, control analyses giving the atomic weight of ordinary lead were carried out in precisely the same way. These yielded essentially the same value as that found by Baxter and Wilson, and more recently, by Baxter and Grover in work as yet unpublished.

The outcome was striking. There can be no question that the radioactive samples contain another element having an atomic weight so much lower than that of ordinary lead as to admit of no explanation through analytical error, and yet so nearly like ordinary lead as not to have been separated from it by any of the rather elaborate processes to which we had subjected the various samples.

All the materials used in the work were purified with the care usually employed in work of this kind. The silver was made by the precipitation of very pure silver nitrate by ammonium formate, and fused upon boats of the purest lime in hydrogen. The hydrochloric acid gas used for fusion and precipitation was obtained by dropping pure sulfuric acid into chemically pure concentrated hydrochloric acid, furnished by a trustworthy firm and known to be very pure. It was carefully dried and freed from spray by many towers of glass pearls, drenched with sulfuric acid. For precipitation this acid was dissolved in pure water in a quartz flask. The water, and also the nitric acid and other substances used in the work, were purified according to the methods usually employed at Harvard for this purpose.

The description of the preparation of the various samples of lead demands further elaboration. [I am omitting that elaboration from this excerpt, though. --CJG]


Sources of Samples of Lead Chloride.

Sample A. Commercial lead acetate, Germany.
Sample B. Carnotite, Colorado, U.S.A. (impure) (Fajans).
Sample C. Carnotite, almost pure.
Sample D. Carnotite the most carefully purified.
Sample E. Commercial lead nitrate, America.
Sample F. Pitchblende, Cornwall, England (Ramsay).
Sample G. Pitchblende, the most carefully purified.
Sample H. Thorianite, Ceylon (Boltwood).
Sample I. Pitchblende, Joachimsthal, Bohemia, purest (Fajans).
Sample K. Pitchblende, preliminary product.
Sample L. Pitchblende, same as I..
Sample M. Thorianite, Ceylon (Miner).
Sample N. Thorianite, later fraction.
Sample O. Uraninite, North Carolina, America (Gleditsch).
Sample P. Extremely careful purification of sample D.
Sample R. Sample O, further purified.


Thus the final analysis yielded results essentially like the preliminary ones. The situation will become clearer if the results are all collected and averaged in a summarized table (V) giving the values of the atomic weight corresponding to each kind of lead.

Table V
Final values found for atomic weight of lead from different sources
Lead from North Carolina uraninite (Sample R)206.40
Lead from Joachimsthal pitchblende (Sample I, K)206.57
Lead from Colorado carnotite (Samples D and P)206.59
Lead from Ceylonese thorianite (Samples H, M)206.82
Lead from English pitchblende (Sample G)206.86
Common lead207.15

The result is amazing. Evidently then the chemical equivalents of these different specimens are markedly different from one another. Because the method of analysis was the same in each case, one cannot help thinking that there is a real variation in the chemical equivalents of these samples of lead. Either a large amount of some element having a chemical equivalent nearly as great as lead, or a small amount of an element having a low chemical equivalent, must be present, mixed with the substance which we ordinarily call lead. The fact that all the analyses were carried out by the same method, and that each sample gives consistent results, seems to exclude the effect of analytical error. The nature of the admixture it would be perhaps premature to decide. Clearly it has reactions very much like those of lead, if not exactly identical; for the various processes to which our material was subjected would have eliminated any element widely different. Moreover, the fact that protracted purification had no effect on the atomic weight of any one sample is evidence in the same direction.


That lead should be composed of a mixture of substances of different origin but similar properties is, after all, possibly not so revolutionary a proposition as might appear at first sight. Rare earths are often very similar in properties, and large amounts of material and very patient fractionation are necessary to separate them. Why should not the same thing be true of several of the commoner elements? The only practical difference besides the presence of radioactivity seems to lie in the fact that in the present case the intruders produce no obvious change in the ultra-violet spectrum. But if all lead is a mixture, this might be expected.

At first sight one might be inclined to feel that the irregularity in the quantitative results above described should diminish one's respect for the significance of atomic weights in general, but further thought shows that this is a superficial view. If the results which we have obtained really indicate that several kinds of lead having the same properties and spectrum may be mixed together and not separated chemically, it is evident that the atomic weight becomes almost the only criterion, except radioactivity, capable of detecting the admixture and tracing the factors to their source. Thus the study of atomic weights is shown to be not less but more significant than it had been before. To emphasize this point, we may perhaps quote two paragraphs, written seven years ago, long before the theory under discussion had been proposed, and when such ideas were of a rather heretical character.

"Are the supposed constant magnitudes to be measured in chemistry really variable...? [ellipsis in original --CJG] If they are thus variable, is it worth while to expend much labor in determining the values which they happen to possess at any one time under one set of conditions?

"The question as to whether or not the supposed constants of physical chemistry are really not constants, but are variable within small limits, is of profound interest and of vital importance to the science of chemistry and to natural philosophy in general. If the latter alternative is true, the circumstances accompanying each possible variation must be determined with the utmost precision in order to detect the ultimate reason for its existence. As Democritus said long ago, 'the word chance is only an expression of human ignorance.' No student of natural science who perceives the dominance of law in the physical universe would be willing to believe that such variations in a fundamental number could be purely accidental. Every variation must have a cause, and that cause must be one of profound effect throughout the physical universe. Thus the idea that the supposed constants may possibly be variable, adds to the interest which one may reasonably take in their accurate determination, and enlarges the possible field of investigation instead of contracting it."[6]

This matter has received not only speculative but also experimental treatment at Harvard. For many years the possibility that samples of a given element from different sources might have different atomic weights had been considered, and investigated, but never before with a positive outcome. In the first investigation of the atomic weight of copper undertaken by one of us as long ago as 1887,[7] samples of copper obtained from Germany and from Lake Superior were found to give precisely the same atomic weight for this element. More recently the question was revived and in 1897, specimens of calcium carbonate were obtained from Vermont, U. S. A., and from Italy, in order to discover whether or not the calcium in these two widely separated localities had the same atomic weight. Not the slightest difference was found between them.[8] Again, in a very elaborate investigation on the atomic weight of sodium,[9] silver was obtained partly from several distinct sources and sodium chloride was obtained partly from several different samples of German rock salt, and partly from the salt pumped from the Solvay Process Company's mines at Syracuse, N. Y. These preparations, differing widely in the steps of manufacture and in geographical source, all yielded essentially the same atomic weights within the limit of error of the process.[10] Yet more recently Baxter and Thorvaldson,[11] with the same possibility in mind, determined the atomic weight of extra-terrestrial iron from the Cumpas meteorite, which gave a result identical with ordinary iron within the limit of error of experimentation. From these researches it would seem probable that even if an unusual eccentricity may be exhibited by lead, most elements do not as a rule differ from any such cause of uncertainty. Baxter and Grover are now engaged in the examination of ordinary lead from different geographical sources. Perhaps this also contains more than one component, as suggested above.

It would perhaps be premature to indulge in further hypothetical reasoning concerning the nature of this extraordinary phenomenon, but the nature of the variation unquestionably points in the direction of the hypothesis of Fajans and of Soddy.

This paper must be looked upon only as a preliminary one. More time, larger amounts of material, and more chemical experimentation are needed in order to be sure that the reactions of the unknown contaminating element and lead are wholly identical. We hope and intend to continue the study, and solve the highly interesting questions which it presents.

We are greatly indebted to the Carnegie Institution of Washington for much of the apparatus and material used in this research.


In this paper a description is given of parallel experiments determining the equivalent weights of various samples of lead chloride obtained from different sources. It was found that all of the radioactive specimens possess a lower atomic weight than ordinary lead, as determined under identical conditions, the deficiency in one case amounting to as much as 0.75 of a unit.

No simple linear quantitative relationship between the exact amount of radioactivity and the atomic weight was found. The radioactivity of the various samples was compared by means of the quantitative electroscope.

The ultraviolet spectrum of a typical radioactive sample was compared with that of ordinary lead, with the help of G. P. Baxter, in a quartz spectrograph. No difference was found between the spectra of the specimens, except for a trace of copper too small to affect the result, and a negligible trace of silver known to have been present. The inference seems to be that radioactive lead contains an admixture of some substance different from ordinary lead, and very difficult to separate from it by chemical means. This substance cannot be identified in the ultraviolet spectrum of the material, either because it has the same spectrum as lead, or because it has no spectrum in that part of the field, or because its spectrum is masked or aborted by that of lead.

This amazing outcome is contrary to Harvard experience with several other elements, notably copper, silver, iron, sodium, and chlorine, each of which seems to give a constant atomic weight, no matter what the geographical source may have been. No attempt is made here to discuss the theoretical aspects of the facts presented, but attention is called to their qualitative agreement with the hypothesis of Fajans and of Soddy.

[All footnotes are from the original publication. --CJG]

[1]Hönigschmid, Monatsh., 33, 253 (1912).

[2]K. Fajans, Ber., 46, 422 (1913); F. Soddy, Chem. News, 107, 97 (1913); see "The Chemistry of the Radio-Elements," Soddy, II, 3 (1914).

[3]Mr. Max E. Lembert, Dipl. Ing., a pupil of Dr. Fajans, was sent by him and the Technische Hochschule of Karlsruhe, with the support of Professor Bredig, to Harvard University especially for this purpose. Sir William Ramsay, also, at about the same time, had urged on behalf of Dr. Soddy that the atomic weight of radioactive lead should be studied in the Wolcott Gibbs Memorial Laboratory. It is needless to say that the opportunity was welcomed; indeed, the matter would have been taken up here before, except for a fear of trespassing upon a field which might properly be considered as belonging to the proposers of the theory. A brief announcement of this work was made by Dr. Fajans at the meeting of the Bunsen Gesellschaft in Leipzig on May 21st, and a brief notice was published in "Science" on June 5, 1914. --T.W.R.

[4]Baxter and Wilson, Proc. Am. Acad., 43, 363 (1907).

[5]Richards, "The Faraday Lecture of 1911," J. Chem. Soc., 99, 1203 (1911).

[6]Richards, (Berlin Inaugural), Science, N. S., 26, 562 (1907); also Die Umschau, 13, 542-543 (1909); translated by F. Haber.

[7]Richards, Proc. Am. Acad., 23, 179-80 (1887).

[8]Richards, THIS JOURNAL, 24, 374 (1902).

[9]Richards and Wells, Carnegie Instit. Wash. Pub., 28 (1905).

[10]Even if both specimens of salt came originally from a Silurian ocean, the time and condition of deposition were probably widely different.

[11]THIS JOURNAL, 33, 337 (1911).

Back to the list of selected historical papers.
Back to the top of Classic Chemistry.