ON ELEMENT 72 (HAFNIUM).

F. Paneth.

1. THE DISCOVERY OF HAFNIUM.

At the beginning of 1923 the situation regarding the question of the element with atomic number 72 was as follows. Dauvillier (Dauviller) \[16\] asserted that, by means of X-ray spectroscopic methods, he had succeeded in establishing the identity of one of the rare earths—namely celtium, described by Urbain (Urbain) \[38, 39\] in 1911—with the element 72 previously absent from the periodic system. This result, however, could not be reconciled with Bohr’s (Bohr) theory of atomic structure, which had been so brilliantly confirmed in all other areas¹). According to Bohr’s theory, element 72 should be not one of the rare earths, but the nearest higher homologue of zirconium. It was precisely for this reason, owing to the prevailing confidence in the correctness of Bohr’s theory, that element 72 was entered in my article on the “Periodic System of the Elements”²) in the table of the elements as a homologue of zirconium, despite Dauvillier’s indications to the contrary; and since every chemist knows well that homologous elements are often found together in nature, that article advised “also seeking element 72 in zirconium minerals” \[30\].

Following the same line of thought, Coster and Hevesy (D. Coster und G. v. Hevesy) carried out an X-ray investigation of zirconium minerals at the Institute for Theoretical Physics in Copenhagen. The result of their work is very important both for physics, where it removes a very substantial and, apparently, experimentally well-founded objection to Bohr’s theory, and for chemistry—as the discovery of a new chemical element.

¹) On this point see especially Niels Bohr’s speech on receiving the Nobel Prize \[7\]: “Repeated investigation has shown that the supposition that the element with ordinal number 72 has the same chemical properties as the rare earths would require a change in the strength of the electron bonds with change of ordinal number, which, apparently, cannot be reconciled with the general requirements of quantum theory.” (p. 624.)

²) “Ergebnisse der exakten Naturwissenschaften”. Erster Band. Berlin. Springer 1923.

As Bohr[^7] had already predicted on December 11, 1922, when receiving the Nobel Prize in Stockholm, Coster and Hevesy were able to give real proof that element 72 is found not among the rare earths, but in zirconium minerals. It should be noted that the very first sample—the Norwegian zircon—which they investigated made it possible to detect appreciable quantities of element 72. The first detailed indications on this subject were given in two letters published in the English journal Nature [^11,^12], and in one letter published in the German journal Naturwissenschaften [^15]. In the subsequent investigations, besides Coster and Hevesy [^13,^14,^24,^30], Hansen (H. M. Hansen) and Werner (S. Werner) [^21,^22] took a special part.

The new element was named by the scientists who discovered it hafnium (symbolic designation \(Hf\)) after the city of Copenhagen (Latinized name—Hafnia), where Bohr’s theory was created.

2. PROPERTIES OF HAFNIUM.

The properties of hafnium correspond completely to the properties of the higher homologue of zirconium; therefore it can be separated without difficulty from the trivalent rare earths, but obtaining it in a form free from zirconium is extremely difficult. Almost all zirconium minerals also contain hafnium, though for the most part in quantities not exceeding 5%. True, immediately after the publication of the first assertions concerning the discovery of zirconium, Goldschmidt and Thomassen (V. M. Goldschmidt und L. Thomassen) [^20] succeeded in finding in Christiania the zirconium mineral alvite, which showed a hafnium content of up to 15%; but such a high hafnium content was found only in individual pieces, while the greater part of the samples contained hafnium in no greater quantity than other zirconium minerals. In the preparation of zirconium compounds, hafnium remains in them as well, and thus the commercial “pure” zirconium salts investigated up to now contain from \(1/2\) to 5% hafnium. The approximate content of hafnium in the earth’s crust is estimated, on the basis of mineral analyses carried out thus far, at 0.001%1. Perhaps the most surprising thing in the history of the discovery of hafnium is that it was discovered only after chemists were compelled to look for it among the com—

corresponding elements under the influence of similar indications obtained from the quantum theory of the structure of atoms.

Although hafnium is found in the fourth group of the periodic system between zirconium and thorium, in its chemical properties it is much closer to zirconium. This is clear already from the fact that typical thorium minerals—and consequently also commercial thorium preparations—do not contain hafnium, whereas the chemically extremely similar elements zirconium and hafnium always remain inseparable both in the process of mineral formation and in the ordinary purification of zirconium-salt preparations. An actual method of separating hafnium from zirconium consists in fractional crystallization of the acid potassium or ammonium fluorohydrogen salt: in the mother alkaline solution hafnium accumulates, in the crystals—zirconium, and it presents no difficulty to obtain by this method zirconium free from an admixture of hafnium; on the contrary, it is much more difficult to obtain hafnium completely free from an admixture of zirconium, and even the purest of the hafnium preparations obtained so far still contain 5% zirconium.

Among other salts that have been investigated, we should mention first of all the extremely sparingly soluble phosphate. Until now, the precipitation of a sediment on introducing sodium phosphate into a strong solution of hydrochloric or sulfuric acid was considered an unambiguous qualitative reaction for zirconium [4]. It has now been possible to show that hafnium phosphate is even less soluble than zirconium phosphate1; the solubility curve of the phosphates of zirconium, hafnium, and thorium therefore shows a minimum value for hafnium. Conversely, in hydrofluoric acid and in alkali fluoride compounds, hafnium dissolves to the same extent as zirconium and thorium. As is evident already from these examples, the chemistry of the new element can draw only the most general conclusions from its position in the periodic system; the details, however, must be studied by means of experimental investigations. Thus, it was established that hafnium salicylate precipitates more readily than zirconium salicylate, whereas with ammonia and sodium thiosulfate zirconium is precipitated to a greater extent than hafnium. The difference in the precipitatability by chloroxyquinoline is very insignificant.

A preliminary determination of the atomic weight gave for hafnium—after introducing a correction for the 5–6% of zirconium contained in the preparations—a value lying between 178.4 and 180.2. Since the hafnium atom is, in round numbers, twice as heavy as the zirconium atom, it is evident that all previous determinations of the atomic weight of zirconium, carried out on preparations containing a known percentage of hafnium, gave values for the atomic weight of zirconium that were too high. Indeed,

experiments with zirconium completely freed from hafnium gave, for zirconium, a noticeably smaller atomic weight than parallel experiments with ordinary zirconium containing hafnium.

In the following table are given those lines whose presence in the optical spectrum of hafnium was established by Hansen and Werner [21]. In this table \(\overset{\circ}{A}\) is the length of the corresponding wave in international Angstrom units, measured from iron lines; \(I\) denotes the relative intensity of the lines in conventional units (the strongest lines are marked with a score of “6”). Some of these lines are identical with weak lines which until now had erroneously been ascribed to zirconium. Zirconium preparations completely free from admixture of hafnium, which were studied at the Copenhagen Institute, did not reveal these lines in the spectrum1.

TABLE I.

The X-ray spectrum of hafnium was also measured after the preparation of highly concentrated hafnium preparations; the position and intensity of all the lines precisely confirmed the place of hafnium in the periodic system as the element with atomic number 72 (see Fig. 2).

Investigation of the X-ray spectrum provided the guiding thread not only in the discovery of hafnium, but also in its chemical purification.

Fig. 1. Optical spectrum of hafnium: according to Hansen and Werner—a) Zr: \(K_2ZrF_6\) on a gas carbon, spark exposure 3 min.; b) Hf: \((NH_4)_2HfF_6\) on a gas carbon, spark exposure 2 min.; c) Fe: iron arc, \(3/4\) min.

For this purpose a special method was developed \([^{10}]\), which makes it possible to determine, with great rapidity and accuracy, the percentage

Fig. 2. X-ray spectrum of hafnium (series \(L\); taken with a crystal of calcareous spar; enlarged two and a half times the natural size).

Fig. 3. Photometric curves of the X-ray spectra of three different fractions of hafnium (the copper lines were obtained from a copper anticathode; tantalum was added as an element for comparison).

content of hafnium in the given preparation. By adding known quantities of the related element tantalum, one can establish at what percentage content of tantalum the characteristic lines of hafnium and

tantalum have the same intensity; the corresponding number thus also gives the percentage content of hafnium. In this way a quantitative analysis of the mineral is carried out without any chemical intervention, and there is no doubt that in the future this process will play a very important role for the rapid testing of other minerals as well, especially when the elements in question, like hafnium, can be separated chemically only with great difficulty.

Fig. 3 shows what unusually accurate results are given by this method of quantitative X-ray analysis. In the present case the material is a zirconium solution with a hafnium content of 4%; from this solution zirconium and hafnium were precipitated in 11 fractions of equal size by means of sodium fluorophosphate. To each fraction 10% tantalum was added, then an X-ray spectrum was taken, and the resulting plates were photometered. Curve I gives the ratios between the intensities of the hafnium and tantalum lines in the first fraction, curve II—in the fourth, and curve III—in the eighth. (What matters is not the absolute height of the peak of the photometric curve, which changes depending on the duration of the exposure, but the relative position of the peaks of the tantalum and hafnium lines on one and the same plate.) On the curves one clearly sees the presence of 4% hafnium in the first fraction and the complete disappearance of this element already in the eighth fraction.

3. THE DISPUTE OVER PRIORITY CONCERNING ELEMENTS 71 AND 72.

It goes without saying that news of the discovery of hafnium attracted attention in the most diverse circles, and not only because it concerned the discovery of a new element—a discovery always very important for a chemist from the theoretical point of view—but also because the new element was immediately found in such quantities, and its position in the periodic system gives grounds for expecting from it properties that make its technical application in many directions quite within the realm of possibility. To this must also be added that the discovery of hafnium represented the first case in which X-ray spectroscopy, which in the hands of Moseley had definitely established the positions of the vacant places in the periodic system, made it possible to fill one of these gaps. And, finally, this discovery confirmed, by a very important and clear example, understandable even to nonspecialists in this field, the high degree of reliability possessed by Bohr’s views on the periodic system1.

Nevertheless, the impression made by the discovery of Coster and Hevesy in Germany cannot even be compared with the excitement it caused in the newspapers of Denmark, France, and England. For, leaving aside the fact that newspaper columns in Germany are now too full of burning political news for scientific interests not to have been pushed into the background—leaving that aside, let us merely point out that the question of the discovery of element 72 gave rise in the countries mentioned to a national dispute over priority. We shall dwell briefly on this point, because the polemic that arose from it shed much objective light on the relations between element 71 and element 72. For better understanding, let us first set down here the most important points from the history of the discovery of both elements \[14\].

In 1878 Marignac isolated from a mineral found near Ytterby (in England) a substance which he recognized as a new rare earth; it received the name “ytterbium” and was regarded as an independent chemical element until 1905, when Auer v. Welsbach, indicating a method of separation, announced the discovery that ytterbium is in reality a mixture of two different earths, for which he subsequently proposed the names “aldebarania” and “cassiopeia” \[1\]. More detailed data on the atomic weights and spectra of these elements he published in 1907, somewhat after Urbain \[37\] had appeared with a similar communication; Urbain was the first to publish tables of the spectral lines of these elements and proposed for them the names “neo-ytterbium” and “lutecium.” In the following years Urbain further purified his preparations and observed a gradual change in the spectra and magnetic properties, which prompted him in 1911 to publish a communication on the discovery of a new element, which he named “celtium” and for which he also published a table of spectral lines \[38\].

with complete distinctness, while considerably later, after the appearance in 1921 of the first editions of Bohr concerning the periodic system—the same conclusion was reached by Bury \[8\]; close to him, Ring \[28, 27\] considered that Urbain’s proposal—that the rare earth celtium possesses atomic number 72—was exceedingly improbable. On the basis of considerations of a different kind, F. Kirchhof \[27a\] regarded the missing element 72 as eka-zirconium. But none of these arguments possessed such a degree of certainty as did Bohr’s investigations, based on the theory of atomic structure. Nevertheless, it is to some extent surprising that, outside Bohr, and in the presence of countless attempts to decompose the rare earths into their components, zirconium was not more often subjected to investigation on the supposition of the possible complexity of its composition. (For a review of the most important works relating here, see the literature references \[14, 24 and 25\].) It is possible that this omission was materially influenced by the circumstance that, in the most widespread presentations of the periodic system, cerium is placed as the higher homologue of zirconium, so that in the titanium, zirconium, cerium, and thorium group there was apparently no place for an unknown element.

ON ELEMENT 72 (HAFNIUM)

Urban’s preparation of celtium was examined roentgenospectroscopically in 1914 by Moseley. Moseley showed that it was a mixture of previously known earths, and that only the appearance of a single new line could be ascribed to some new element—celtium \([29]\). But since then the technique of X-ray investigations—thanks to the work of Siegbahn and his pupils—has made such great progress that Dauvillier’s indication of two “exceedingly weak” X-ray lines, found by him in the investigation of that very same preparation and belonging to element 72, was at first accepted by specialists as reliable \([32]\). Only Coster and Hevesy showed that the lines belonging to element 72 proceed not at all from any of the rare earths, but from the quadrivalent homologue of zirconium, the element “hafnium.”

Against this name, just as against the opinion that the Copenhagen investigators had discovered a new element, Urban came forward in a series of articles, and the French daily press agreed with him in its appraisal of the reports of Coster and Hevesy. First of all Urban attempted to defend his assertion that the celtium he had found really is element 72, that it was obtained by him in a fraction with the rare earths; the merit of Coster and Hevesy, however, consists in the fact that in zirconium minerals they found a richer source of this new element \([41]\). In due course, as proof of the existence of the new element celtium, Urban published its optical spectrum; but an investigation of the optical spectrum of hafnium carried out by Hansen and Werner \([27]\) at Bohr’s institute found that the spectrum of hafnium does not contain the slightest traces of any of the lines of Urban’s celtium. This refuted the assertion of the identity of celtium and hafnium. Urban then adopted \([40]\) a somewhat different point of view. While admitting the possibility that his report of the discovery of celtium in 1911 was based on an error, he nevertheless ascribed to himself and Dauvillier the honor of discovering element 72 in 1922, since they had been the first to observe X-ray lines belonging to element 72.

At first this question represented only a limited, purely personal, and by no means scientific interest, since the observation of two exceedingly weak lines, which had been erroneously assigned to a non-existent rare earth—and moreover their observation in a region where they could have been theoretically expected—not only did not in the least broaden our natural-scientific knowledge, but even constituted an obstacle on the path toward the search for element 72 outside the group of rare earths. But even this claim to the honor of the first observation of the X-ray lines of element 72 Dauvillier and Urban were unable to defend. Urban at one time asserted that his preparations had been completely purified of all impurities not belonging to the rare earths, and that therefore

in their spectra one cannot find, for example, any zirconium lines. But, as Coster and Hevesy showed, from the absence of zirconium there necessarily follows also the absence of hafnium, since both these elements are precipitated to the same degree by oxalic acid. With this there is in full agreement the circumstance that, as the investigations of Hansen and Werner showed, the spectrum published by Urbain contains not a single hafnium line. Proceeding from this, the Copenhagen investigators were able to show that in the X-ray spectrum, too, both “extremely weak” lines, which are now the sole proof in favor of the French claims, differ by 4 X-units \((=4\cdot10^{-11}\ \mathrm{cm}\), which, according to Dauvillier’s data, corresponds on his photographic plates to more than \(0.6\ \mathrm{mm})\), i.e. by an amount greater than the limit of possible error, from the hafnium lines measured by Coster and Hevesy. On the other hand, it is not even clear how Dauvillier could have observed the line \(L\beta_2\) of hafnium, whereas the stronger line \(L\beta_1\) did not appear on his plates. It is likewise difficult to understand how he succeeded in observing the line \(L\beta_2\), when the still stronger line \(L\alpha\) lies at the limit of visibility. Therefore one must consider—unless one admits the possibility of self-deception in the case of “extremely weak” lines, suggested by a definite expectation—that the disputed lines are lines of higher order arising from some impurities in the preparation, but by no means lines of hafnium.

As a conclusion to this dispute over the question of priority, a genuinely important positive truth was found: namely, when the experimenters of Bohr’s institute, after doing all the work, turned to the question of how it could have happened that so experienced a rare-earths scientist as Urbain, on the basis of the optical spectrum and magnetic susceptibility, believed that he held in his hands a new earth for which, after the discovery of hafnium, there was in general no place in the periodic system. Hansen and Werner \[22\] were able to show that the optical spectral lines which Urbain published for the supposed “celtium” are nothing other than lines of element 71 (“cassiopeium,” according to Auer, or “lutecium,” according to Urbain), and that—as they likewise showed after a careful review of the literature—the greater part of Urbain’s celtium lines had been measured in 1915 by Eder (Eder) \[18\] in Vienna, and some even in 1911 by Exner and Haschek \[19\] on Auer’s preparation of cassiopeium. As they showed further, these lines appear only in the case when the preparation is sufficiently purified, and in this case acquire sharpness only when placed on the anode of a voltaic arc. Therefore Urbain’s results must be understood to mean that in 1907, when he first described “lutecium,” his preparation contained only a very small quantity of this element 71, and only after many years of further concentration did he

been content to observe the diffuse lines which, in 1911, he ascribed to the presence of a new rare earth, then called by him “celtium.”

This supposition acquired full certainty after Coster and Hevesy [14] showed that Urbain’s data on the magnetic susceptibility of the proposed celtium were in full agreement with their data. The circumstance that the paramagnetism of his preparations in 1911 was three or four times less than before is a necessary consequence of the supposition of a gradual increase in the content of element 71 in Urbain’s preparations. For Stefan Meyer [28], as early as 1908, measured on Auer’s cassiopeium preparation just as small a value of paramagnetism as that which Urbain found in 1911 for his substance in which, in his own opinion, the greatest quantity of the hypothetical element “celtium” was contained.¹ Thus these optical and magnetic investigations on element 71 lead to the conclusion that not only did Auer von Welsbach give the first indications of the decomposition of ytterbium two years earlier than Urbain, but that at the same time his preparations were far more concentrated than Urbain’s preparations, which contained only an extremely insignificant quantity of element 71. In the light of these new investigations there is no basis whatever for using, instead of the name proposed by Auer, the name proposed by Urbain, which was at one time adopted by an international commission consisting of: Clarke, Ostwald, Thorpe, Urbain [3]. Therefore, while preserving in the following tables the historical name “ytterbium” for element 70, we, together with the Copenhagen scholars, designate element 71 after Auer as “cassiopeium” (Cp)²; the element “celtium,” representing concentrated cassiopeium, must be erased from the literature.

The second dispute—concerning priority in the discovery of hafnium, which arose in England—was resolved much more quickly. In 1913 A. Scott investigated a titanium-bearing ferruginous sand from New Zealand, leaving a small, sparingly soluble residue unanalyzed. On receiving the first news of the discovery of hafnium, Scott, on the basis of an approximate determination, took his residue to be hafnium oxide and proposed to name element 72 not hafnium but “oceanium,” on the grounds that his preparation came from Oceania and also because of its closeness to titanium [33].

¹ The insignificant magnitude of the paramagnetism of element 71 also agrees perfectly with Bohr’s theory, since at this place in the periodic system the formation of tetravalent electron groups must already be completed.

² It should be mentioned that even in earlier years in Vienna attempts were made [3] to restore Auer’s indisputable priority for the naming of this element. But it was probably only the new facts discovered in Copenhagen and speaking in the same sense that made a reconsideration of this question necessary.

English and Danish newspapers racked their brains over the question of who should christen the “new element,” and published telegrams, interviews, and articles on this subject. Meanwhile, Bohr’s institute established that the sample of “oceania” sent by Scott did not give a trace of a single X-ray line of element 72, and Scott himself, upon reanalysis, found that his residue was simply titanium dioxide, in which part of the titanium had been replaced by silicon. He, therefore, without a moment’s hesitation, withdrew his claims to the discovery of element 72 \[34, 35\]. It would, strictly speaking, be superfluous to dwell on this, had the preceding example not shown that disputes about priority are sometimes carried on in science with astonishing persistence1.

4. THE PERIODIC SYSTEM AFTER THE DISCOVERY OF HAFNIUM.

In view of the very great significance that the discovery of hafnium, which delimited the group of the rare earths, has for the formulation of the entire periodic system, we consider it not superfluous to give, as an appendix, two tables of the periodic system, which give an account of the present state of our knowledge. They are, of course, in the closest connection with the scheme given by Bohr \[6, 7, 30\]; but, in the interests of the teaching of chemistry and of practically working chemists, we thought it useful to give, alongside this representation—which is irreplaceable for theoretical investigations—also several simple tabular arrangements of the elements, somewhat modernized from the point of view of Bohr’s theory and the discovery of hafnium, such as are still used in all chemistry textbooks.

Among the most widespread representations of the periodic system we distinguish the “table with long periods,” recommended by A. Werner (A. Werner)\[42\] and many others, and the still more widespread “table with short periods” \[31\], each of which has its advantages for teaching. The table with long periods (Table II) is obtained directly from Bohr’s scheme by appropriate simplifications. But even in the table with short periods (Table III) we have tried not to deviate from Bohr’s investigations, insofar as this is possible while wishing to preserve a simple and, as far as possible, easily surveyed tabular form. Thus, for example, lithium, sodium and, in exactly the same way, beryllium and magnesium are placed between subgroups a and b of the first and second groups, because they are connected with both subgroups, according to the investigations of the theory of atoms; whereas the initial members of groups from the third to the seventh belong only to subgroup b. The atomic weights are taken from the list of the German Atomic Commission for 1923 \[17\].

Everything further is self-evident upon inspection of the table.

TABLE II.

GROUPS.
(Mendeleev’s table)

*) Rare-earths

TABLE III.

*) Rare earths

Literature.

1) Auer v. Welsbach, C. Wiener Anzeiger 1905, Bd. 10; Wiener Sitzber. 1906, Bd. 115; Lieb. Ann. 1907, Bd. 351, pp. 458, 464.
2) — Wiener Sitzber. 1907, Bd. 116.
3) — Monatshefte f. Chem. 1908. Bd. 29, pp. 181, 204.
3a) Bardet, I. Comptes rendus 1923. T. 176, p. 1711.
4) Biltz, W. and Meklenburg, W. Zeitschrift f. angew. Chem. 1912, Bd. 25, p. 2110.
5) Bohr, N. Zeitschr. f. Physik 1922, Bd. 9, p. 1.
6) — Drei Aufsätze über Spektren und Atombau. (Publ. Vieweg, Braunschweig 1922), p. 70.
7) — Über den Bau der Atome (speech on receiving the Nobel Prize). Die Naturwissenschaften 1923, Bd. 11, p. 606. (Translation in Uspekhi Fizicheskikh Nauk, vol. III, issue 4).
8) Bury, C. R. Journ. Amer. Chem. Soc. 1921, Vl. 43, p. 1602.
9) Clarke, F. W., Ostwald, W., Thorpe, T. E. and Urbain, G. Ber. d. Deutschen Chem. Ges. 1909, Bd. 42, p. 11.
10) Coster, D. Speech in the Bunsen Society, Hanover 1923 (published in Zeitschr. f. Electrochem. 1923); Chem. News 1923, Vl. 127, p. 65.
11) Coster, D. and Hewesy, G. v. Nature 1923, Vl. 111, 20 Jan. 1923.
12) — Ibidem, 10 Feb. 1923.
13) — Ibidem, 24 Feb. 1923.
14) — Ibidem, 7 April 1923.
15) — Naturwissenschaften, 1923, Bd. 11, p. 133.
16) — Dauvillier, A. Comptes rend. 1922, T. 174, p. 1347. 1923.
17) German Commission for the Determination of Atomic Weight. B. d. Deutsch. Chem. Gesellschaft, 1923, Bd. 56, special supplement to the April issue; Zeitschr. f. phys. Chem. 1923, Bd. 105, p. 1.
18) Eder, J. M. Wiener Sitzber. 1915, Bd. 124.
19) Exner, F. and Haschek, E. Spektraltafeln.
20) Goldschmidt, V. M. and Thomassen, L. Norsk Geologisk Tidsskrift 1923, Bd. 7, Heft 1.
21) Hansen, H. M. and Werner, S. Nature 1923, Vl. 111, 10 March.
22) — Ibidem, 7 April.
23) Harkins, W. D. Journ. Amer. Chem. Soc. 1917, Vl. 39, pp. 856, 867.
24) Hevesy, G. v. B. d. Deutsch. Chem. Ges. 1923, Vl. 56, p. 1503.
25) — Ibidem, 7 April.
26) King, H. S. Journ. Amer. Chem. Soc. 1922, Vl. 44, p. 323.
27) — Nature 1923, Vl. 112, p. 9.
27a) Kirchhof, F. Zeitschr. f. phys. Chem. 1920, Bd. 94, p. 253; Physikal. Z. 1920, Bd. 21, p. 711 ff.; Österr. Chemiker-Zeit. 1923, Bd. 26, p. 119.
28) Meyer, St. Wiener Sitzber. 1908, Bd. 117, p. 455.
29) Moseley, H. G. J. Nature 1914, Vl. 94, p. 353.
30) Paneth, F. Das periodische System der chemischen Elemente (in the book “Ergebnisse der exacten Naturwissenschaften,” I. Springer, Berlin 1922. Bd. I, pp. 362 to 405. See especially pp. 383 and 398).
31) — Zeitschr. f. angewandte Chem. 1923.
32) Rutherford, F. Nature, 1922, Vl. 109, p. 781.
33) Scott, A. Journ. Chem. Soc. 1923, Vl. 123, 124, p. 311.
34) — Ibidem, p. 881.

35) Scott, A. Nature 1923, Vol. 109, p. 781.
36) Thomsen, Julius, Zeitschr. f. anorg. Chem. 1895, Bd. 9, p. 190.
37) Urbain, G. Comptes rendus 1907, Vol. 145, p. 759.
38) — Ibid., 1911, Vol. 152, p. 141.
39) — Ibid., 1922, Vol. 174, p. 1349.
40) — Ibid., 1923, Vol. 176, p. 496.
41) Urbain, G. and Dauvillier, A. Nature 1923, Vol. 111, p. 218.
42) Werner, A. Ber. d. Deutsch. chem. Ges. 1905, Bd. 38, p. 914. Neuere Anschauungen auf dem Gebiet der anorganischen Chemie (Vieweg, Braunschweig).1

Translated by L. Timerman.

1. Paneth’s article was originally printed in the collection Ergebnisse der exakten Naturwissenschaften. Zweiter Band. — Berlin 1923. J. Springer, p. 163. — Ed. ↩↩↩↩↩↩