ties in a Chain of Atoms, by ARTHUR LAPWORTH and ROBERT ROBINSON.

resem

a en

In

The deductions of the Principle of Induced Alternate Polarities put forward by Lapworth and by Kermack and Robinson, though having little superficial blance in some respects, are, on critical examination, found to be identical in others, and suggest very similar applications of the rule. The range of phenomena embraced by the theories is very wide, but a complete exposition of the subject in relation to reactivity, orientation, mechanism. of reactions, intramolecular rearrangements, tautomerism, co-ordination, triphenylmethyl, and physical properties of substances could only be attempted in monograph, and such a work we are gaged in preparing for publication. broad outline the principle may be said to have developed as a common expression of such well-known rules as those of Markownikoff, Michael, and Crum Brown; in particular, the regularities observed in the course of substitutions in aromatic compounds had a specially noteworthy importance in revealing the wider generalisation. It is therefore not a little surprising to find that some recent contributions to the subject reverse the order, and, employing the general theory as a basis, proceed to point out that certain groups of facts are in h"mony with the principle. For example, Fraser and Humphries (The Chemical News, 1923, CXXVI., 257) find that the theories may be applied to the manifold problems of orientation in the benzene series.

Lowry's theory of mixed double bonds. fails to suggest any explanation whatever of the most important facts which led to the recognition of the "Principle of Induced Alternate Polarities, a few of which facts have been mentioned above, and it inadmissible as an explanation of the Principle, since it introduces theoretical limitations where no distinctions are in practice to be perceived. There are theoretical diffculties also. The remark that a double bond in organic chemistry usually reacts as if it contained one co-valency and one electrovalency is, with its double qualification, unexceptionable, being no more than a restatement of the fact that unsaturated compounds undergo polar additive reactions.

Octet Stability in Relation to Orientation and Reactivity in Carbon Compounds, by ROBERT ROBINSON.

The explanation of the alternating polar effect advanced by Kermack and Robinson has apparently been misunderstood in some quarters, and further remarks on the subject may not be out of place. Two deductions from the octet theory were made, and the second was an interpretation of the mechanism of addition to conjugated systems, a precise translation in electronic terms of earlier representations by means of partial valency symbolism. Here it was shown that the a-y-rule is in part a consequence of the preservation of octets in the intermediate atoms of the chain, and the schemes given are fundamentally identical with some of the cases presented by Lapworth in a more general form. Some time before, the writer had suggested that the residual or additional partial valencies of Thiele should be replaced by real partial valencies, which in the sum are equivalent to the normal valency, and the interest of the translation in electronic terms partly resided in the fact that it was possible only when this real partial valency theory was adopted. So natural does this suggestion appear that in several recent text-books and memoirs the theory is ascribed to Thiele, although this author did not adopt a partial valency symbol derived by splitting a normal valency, and his benzene formula was not, as many writers now assume, more symmetrical than that of Kekulé.

In connection with the first deduction from the octet theory, which concerns what may be called static alternate polarity, misconception has arisen. The octet theory was employed in a more or less symbolic fashion, making no assumptions in regard to the position of electrons or the precise significance of covalency. The views advanced are considered to be just as sound on the basis of the Rutherford-Bohr atom

as on the Lewis-Langmuir atom. The essential point is that there are certain stable electronic configurations associated with individual atoms and termed for convenience octets; although if they are duplets, sextets, decets, or dodecets the argument is unaffected. The second stage is that there must be some octet-stability factor which is not wholly the outcome of the equilibrium of electrostatic forces. It should be remarked that the words stable and unstable are used in reference to behaviour in the course of reactions.

THE SEPARATION AND DETERMINATION OF SODIUM AND LITHIUM BY PRECIPITATION FROM ALCOHOLIC PERCHLORATE SOLUTION.' BY H. H. WILLARD AND G. FREDERICK SMITH.

[Contribution from the Chemical Laboratories of the University of Michigan and the University of Illinois.]

A review of the literature on the separation of sodium and lithium shows that none of the methods so far suggested is entirely satisfactory. The solubility corrections are usually large and the separation must be repeated one or more times.

Most of the methods proposed involve the extraction of lithium chloride from the anyhdrous mixed chlorides of the metals involved, using an organic solvent or mix. ture of solvents in which the chlorides other than lithium chloride are insoluble. Another general type of procedure consists in dissolving the mixed chlorides in the least quantity of water necessary for their solution, followed by the precipitation of the chlorides other than lithium chloride, by the addition of an organic solvent or mixture of solvents in which the chlorides precipitated are but slightly soluble, a correction being applied for this slight solubility.

A full and critical discussion of the papers relating to this subject up to the year 1912 is given by Skinner and Collins.2 Of the more recent work, that of Winkler3 employs isobutyl alcohol in an extraction emthod, more than one extraction being required in each analysis. The method is tedious in operation and subject to errors, because of the number of manipulations involved. In the method of Palkin' a concentrated aqueous solution of the mixed chlorides is precipitated by the addition of anhydrous alcohol followed by ether.

The

chlorides thus precipitated are filtered and the small amount remaining in solution is recovered by evaporating to dryness and extracting with alcohol and ether. The method is an improvement over the others, since it eliminates the solubility correction, but the use of the volatile ether is a disadvantage.

The methods in which chlorides are precipitated rather than extracted most closely approximate the usual analytical processes and are theoretically more accurate. However, Palkin's process is, strictly speaking, not really such, as the precipitant does not carry a component which enters into the composition of the precipitate formed; moreover, the method is applied to saturated rather than dilute solutions.

The object of the present paper is the development of a method for the separation and determination of lithium and sodium, based upon precipitation of the material separated, in a manner analogous to the usual analytical practice.

THE METHOD.

The process, in brief, consists in the precipitation and separation of sodium chloride from a solution of the mixed perchlorates of sodium and lithium in n-butyl alcohol by the addition of a butyl alcohol solution of hydrogen chloride according to the reaction, NaCIO + HCl = NaCl + HCIO. The reagent is added to the solution of the perchlorates until a 6 per cent. acid concentration is attained. The precipitated sodium chloride is filtered on a weighed Gooch crucible, washed with a 6-7 per cent. solution of hydrogen chloride in butyl alcohol, dried at 250° and ignited for a few 'minutes at 600°. The lithium chloride, after removal of the organic matter by evaporation, is determined by conversion to lithium sulphate, a correction being applied for the almost negligible amount of sodium chloride remaining in the filtrate. In some cases the reagents give a slight blank. Potassium cannot be present since its perchlorate is insoluble in alcohol.

PREPARATION OF MATERIALS.

Normal Butyl Alcohol. This material is. readily obtainable on the market at the present time at a moderate price. The alcohol used in this research had a boiling range of 112-118° and a density of 0.8065 at 25°/4° 60 per cent. of this product boiled within a range of 1° of the true boiling point. This fraction, when dried by reflux

ing with a slight excess of metallic calcium, periments were also carried out using alcohad a boiling range of 116.2-116.7° and a density of 0.8060 at 25°/4°. Half of this fraction boiled within a range of 0.05°. In most of the work the 60 per cent. fraction, obtained as described above, was used. Exhol of widely different constants, the data concerning which will be recorded later.

Perchloric Acid.-The method of Willard' was used for the preparation of the perchloric acid. It was twice distilled under a pressure of from 5 to 15 mm. and contained about 72 per cent. HCIO. Ten g. of this material when evaporated in a platinum crucible gave an almost unweighable residue after ignition.

Sodium Chloride.-An imported product of highest purity was used.

Sodium Perchlorate.-This material was prepared by treatment of the purest sodium carbonate with a slight excess of dil. perchloric acid. The anhydrous sodium perchlorate was obtained by crystallisation above 50°, using centrifugal drainage. The product thus obtained was dried in a current of dry air at a temperatrue of 250°. To obtain samples of sodium perchlorate for the analytical separations, pure sodium

chloride in weighed portions was evaporated on a hot plate with a slight excess of perchloric acid until fumes of perchloric acid were no longer evolved.

was

Lithium Chloride.--This material prepared from a product which contained some sodium chloride. It was freed from the latter by solution in hot butyl alcohol under a reflux condenser. The solution was cooled and the sodium chloride filtered off. The lithium chloride in the filtrate was recovered by evaporation in a platinum dish to a party mass which was dried in an electric oven at 100°. It was further heated in a muffle at 500°, and finally fused in a current of hydrogen chloride dried with sulphuric acid. The product thus obtained was cooled, crushed, and powdered under conditions which insured no contact with atmospheric moisture. It was used only in the solubility determinations.

5 Willard, Jour. Amer. Chem. Soc., 1912, XXXIV., 1480.

CORRESPONDENCE.

ISOTOPES OF COPPER.

To the Editors of THE CHEMICAL NEWS. GENTLEMEN, Readers of this Journal may remember that I made an attempt to prove that copper was composed of wholenumber isotopes in such proportion as to give a mean value as the atomic weight of this element, viz., 63.57. The values sug

gested in this Journal of March 20, 1914, page 143, were 63 and 67.

W. D. Harkins, in The Journal of the American Chemical Society for June, 1923, on page 1429, predicts the values 63 and 65.

Now, A. J. Dempster, in Nature of July 7, 1923, records having obtained positive rays of copper by means of a molybdenum furnace heated with a coil composed of molybdenum wire embedded in alundum cement. Three isotopes were observed, having provisionally the values 62, 64, 66.

Should Dempster's values be confirmed by further experiments, it will show the utter futility of attempting precise predictions based upon rather meagre evidence, but the general result indicated as far back as 1914 will have been confirmed.-Yours, &c., F. H. LORING.

NOTICES OF BOOKS.

Atomic Structure and Spectral Lines, by A. SOMMERFELD, Translated from the third German Edition (1922) by H. L. BROSE. Pp. XIII. + 626 and 125 Figures in text. London: Messrs. Methuen & Co., Ltd., 36, Essex Street, W.C.2. 1923. 32s. net.

It is one thing to possess a book and another thing to understand all that is in it. Prof. Sommerfeld's book, now a classic, contains much that requires very special study to understand fully, but there is a great deal of information clearly presented, thanks to the translator; and all who are interested in the fundamental characteristics of the atom, as studied and investigated more particularly by the physicist, should add this volume to their library, and study it.

photo-electric effect and its converse; glimpses of the quantum hypothesis; radioactivity; nuclear charge and atomic number; the atom as a planetary system; laws of radio-active displacement and the theory of isotopes; peripheral and central properties of the atom; visible and Röntgen spectra; configurations of the inert gases; Laue's discovery; survey of the K, L, and M series; corresponding limits of excitation; K series and its bearing on the periodic system; L and M series; doublet relationships; introduction to the quantum theory; empirical data about the spectra of hydrogen; principle of combination; Bohr's theory of the Balmer series; elliptic orbits in the case of hydrogen; quantising the spatial position of Kepler orbits; theory of the magneton; spherical wave and its propagation; conservation of energy and momentum; principle of selection and rule of polarisation; orbits of hydrogen in the Stark effect; Zeeman series effect; quantum theory of the scheme; principle of selection for the azimuthal quantum; testing the series scheme by the method of electronic impact; spectroscopic law of displacement and law of exchange; visible bands; meaning of the head of the band; law of the edge of the band; many lines spectra; preliminaries concerning the theory of relativity; variability of mass and inertia of energy; relativistic Kepler motion; fine structure and the relativity correction; doublets of the Röntgen spectra; spectroscopic confirmation of the theory of relativity, Bohr's principle of correspondence, &c.

Prof. Sommerfeld, as we all know, has developed a theory of the fine structure based upon Bohr's classical work in explanation of the spectra of hydrogen and helium, but the theory in some respects extends to all elements. In the "theory of the fine structures there is a confluence of the three main currents of modern research in theoretical physics, namely, the theory of electrons, the theory of quanta, and the theory of relativity. This is exhibited in a particularly vivid way in the way our finestructure constant is built up :—a 2е he. Here e is the representative of the theory of electrons, h is the worthy representative of the quantum theory, and c comes from the theory of relativity and, indeed, characterises it in comparison with the classical theory."-Page 525.

a

=

From the foregoing it will be seen that in developing a theory of the atom based upon electron movements, much of modern phy

sics has been brought into the problem, and on this account the book becomes a treatise of considerable scope.

Apart from the more rigorous treatment, there are ideas introduced in a suggestive way which will afford food for those in search of advance information. As an example, the question of two kinds of electricity is not made so evident by current electricity, which is a flow of electrons, for in terms of electrons the unity idea is prominent. In atomic physics, however, a positive charge signifies more than the absence of a negative charge, since" positive electricity is always associated with ordinary matter." The difference, therefore, imtypes of electricity that differ not only in sign but also in nature," which are individualised in the electron and the positively charged H atom. Quoting from page 22, We can picture an atom (or a body) as highly charged negatively as we like, that is, we can add to it any number of negative electrons'; whereas, we can only increase the positive charge to a certain maximum amount so long as we do not considerably alter the mass. We can remove from the

plies

46

66

atom only as many electrons as it possesses at the outset. According to the theory of relativity, no inherent or relative change in charge can take place. Electricity thus becomes a substance, but there are two such substances. The charge and the mass are hereby indissolubly associated with one another, the negative charge with the electronic mass, the positive charge with the hydrogen mass."

In discussing polarisation, Prof. Sommerfeld states that "it signifies that a ray favours a certain plane passing through it more than the one perpendicular to this plane." The early idea that Röntgen rays were longitudinal vibrations involves the scheme of symmetry, and therefore polarisation becomes impossible, since Barkla discovered that primary Röntgen rays are partly polarised, and that secondary Röntgen rays are wholly polarised in certain directions. The longitudinal vibration theory thus breaks down, and we are left with the interpretation that the direction of vibration does not correspond with the direction of motion of matter, but that the vibration is that of electric force which participates in the wave-radiation." In wireless telegraphy the emission is zero in the direction of the alternating current, corresponding to its component of acceleration, but at right

angles to the antenna the emission becomes a maximum.

The theory of fine-structure, which Prof. Sommerfeld has developed with great mathematical skill, follows when the Bohr theory of stationary states involving the quantum theory is analysed by means of relativity mechanics involving the change of mass of the electron with a change in its orbital velocity, so that those lines which were regarded as coincident become separated into a configuration of closely associated lines owing, as is evident, to the relativity effect involved. The velocity of the electron is less at the aphelion than at the perihelion, and this difference becomes greater as the eccentricity of the orbit increases. Quoting from page 475: "Thus the observation of the fine-structures discloses the whole mechanism of the intraatomic motions as far as the motion of the perihelion of the elliptic orbits. The complex of facts contained in the fine-structures has just the same importance for the special theory of relativity, and for the atomic structure as the motion of Mercury's perihelion for the general theory of relativity." It is important to note that without the theory of relativity it would have been. impossible to develop the theory of fine

structure.

On p.

In a work of this magnitude errors are bound to occur in the first translated edition. These do not, however, detract from the value of the book. 12, 5th line from top, captivity would be a better word than "capacity." Two lines further down, on the same page, the wording, "Although they here also soon," could be improved. On p. 23, the expression,

soup-plate," seems to imply a shape not strictly accurate, and it is not in dignified keeping with the subject.

On p. 84 it is stated that "isotopic elements cannot be separated from one another by chemical means at all, and exhibit identical physical properties throughout." Atomic mass is a physical property and, in the case of some isotopes of a given element, it varies to the extent of eight units, as stated in the same paragraph. The above sentence is faulty in two respects; for (1), there are some physical properties which are not identical, and (2) the word element is more correctly used in a collective sense, and it should not be used synonomously with the word atom in this connection. An element is a collection of atoms of the same chemical kind with one common atomic number, except in the case of radio-active