of positive and negative charges on alternate atoms. This view differs fundamentally from Thiele's original conception, in that, whereas he chooses maleic acid as a typical example of conjugation, the new theory shows in this compound a discordant series of polarities, which destroys the conjugation of the system. On this point a definite experimental test of the theory can be made, since, if Thiele's view were correct, the maleic type of structure should be particularly stable, whilst in the light of the present theory the crossed polarities would render it relatively unstable, at least under conditions which tend to promote polarization of the molecule. I have already directed attention to three cases in which derivatives of maleic acid get rid of this condition of crossed polarities by undergoing isomeric change to a compound in which the series of alternate single and double bonds is shorter, but in which the polarities are no longer discordant; I also quoted one case † in which the crossed polarities are rendered concordant by the loss of a molecule of carbon dioxide. In another group of cases, which has been studied intensively by Thorpe and his colleagues ‡, the crossed polarities are eliminated by the addition of water or of ethyl alcohol. In the absence of any alternative explanation of these changes, and especially in view of the fact that Thiele's views lead to a directly opposite conclusion, the series of cases quoted above may be regarded as decisive evidence in favour of the theory of polar double bonds. (b) It is also of interest to notice the conditions under which these changes occur. Loss of carbon dioxide and gain of water and alcohol actually occur during the course of reactions carried out with a view to preparing acids of a maleic type; it is therefore likely that these changes take place whilst the molecule is in the reactive or polar form. When, however, a compound of a maleic type has been isolated, the isomeric changes that have been referred to above are brought about by boiling with strong caustic soda, i. e. under the very conditions that have been postulated (p. 970) as likely to drive a molecule over into a polar form. Opposite conditions can be achieved by converting the acid. into its anhydride, since this not merely removes the sodium ions of the salt but also eliminates even the acidic hydrogen of the acid. All possibility of forming free ions is thus removed and, in the absence of these ionizable key-atoms, it is * Trans. Chem. Soc. cxxiii. P. 825 (1923). + Ibid. Trans. Chem. Soc. cxxi. p. 650 (1922). doubtful whether even the polarity of the carbonyl groups would be maintained. It is precisely under these conditions that a double bond, which has been expelled from its original position under the influence of caustic soda, creeps back again into the ring and yields a derivative of maleic anhydride. (e) A number of other cases are known in which strong caustic soda is used to bring about chemical or physical changes, e. g. a reversal of sign in the rotatory power of the tartrates. It will be of interest to inquire whether in these cases also the chemical changes demand the formation of a polarized molecule as an intermediate or final product, and whether the physical properties developed by the soda are those that might be expected in a molecule of this kind. (In a totally different field crossed polarities afford a simple explanation of the mobility of aromatic groups which are "incorrectly" oriented, e. g. the meta nitro-group of B and trinitrotoluene. Apart from the theory of induced polarity, it is difficult to see any clear reason why nitrogroups occupying adjacent positions in the benzene ring should be so mutually antagonistic. 7. Multipolar Ions. An important application of the theory of mixed double bonds is found in the discovery of the multipolar ion. i. e. of ions which, in addition to the electrification which is required to give the net charge of the ion, possess a series of additional positive and negative charges, so that, for instance, a univalent anion may contain three negatively charged atoms, alternating with two positively charged atoms, giving the net charge of minus one. This discovery has provided a novel interpretation of the phenomenon of tantonierism, and has disclosed a relationship between tautomerism and coordination which appears not to have been recognized previously, even by those who were working on these two phenomena separately, since no definite liaison had been established between these two independent lines of investigation. (a) Tautomeric Ions.-It has usually been assumed that ethylacetoacetate can yield two ions, one derived from the ketonic form of the ester, and the other from the enolie form; and that the reversible isomeric change of the esters (which appears to demand the presence of a trace of an alkaline catalyst) involves both the migration of a mobile hydrogen ion and a structural rearrangement of the anion. If, however, the views set out in paragraph 4 (b) above are adopted, it would appear that, in the sodium derivative at least, the metallic atom may act as the key-atom in setting up "induced alternating polarities" throughout the conjugated chain, as shown in the following scheme : - + CH. CONA: CH. CO. OEt→→→ NaO.CMe.CH.C(OEt).O. The ion of the ester is now seen to contain three negative poles. It is tautomeric, since if a hydrogen ion be attached to the negative pole 2 it will yield a ketone, whilst if attached to the negative pole 1 it will yield an enol having the formula conventionally assigned to this compound; if attached to the negative pole 3 it would yield an isomeric enol, in which the carboxyl-group instead of the ketonic-group has been enolized. No evidence of the existence of a second enol has been discovered, and of this fact also the electronic theory provides a simple explanation, which I have discussed in a paper on "Coordinated Hydrogen" recently communicated to the Chemical Societyt. This explanation is based on the assumption that in the sodium-derivative, a sodium ion is coordinated with two oxygen poles 1 and 3 in just the same way as in the metallic derivatives of acetylacetone ; and that in the enol a hydrogen ion may be similarly coordinated. Since the distinction between principal and subsidiary valencies cannot be maintained, the coordinated forms of the two enols would be identical. This coordinated enol may be expected, like the sodium derivative, to persist in a polarized form, since it owes its very existence as a stable isomer to the possibility of forming by coordination a conjugated six-atom ring; but even this apparently obvious conclusion ought to be confirmed by experiment. The ketonic form, on the other hand, like maleic anhydride, contains only two unconjugated carbonylgroups, and may easily revert to a non-polar form. Indeed, it is possible that di-ketones of this type give rise in general to one polarized and one unpolarized form, and that these are represented by a fully-polar coordinated enolic and a non-polar ketonic form respectively. *In the second formula the groups have been re-arranged so as to show the conjugated chain as the principal axis of the molecule. Trans. Chem Soc. cxxiii. p. 2111 (1923). Morgan, Trans. Chem. Soc. cv. p. 193 (1914). (b) Coordination of Multipolar Ions.-The importance of multipolar ions in the preparation of coordinated compounds has been pointed out in the Chemical Society paper already referred to, and has been discussed more fully in a series of articles in Chemistry and Industry.' It will therefore suffice here to direct attention to the fact that we have already succeeded in preparing in this laboratory coordinatedderivatives of compounds which have been studied hitherto as examples of dynamic isomerism, and that the liaison between these two lines of investigation has provided another very fertile field of research. 8. Summary. (a) The theory of intramolecular ionization can be extended to organic compounds if it be assumed that double bonds can assume a form in which one carbon atom carries eight L-electrons, but the other only six, one pair of electrons only being shared. Since the former atom is then negatively and the latter positively charged, this type of double bond may be represented as made up of one covalency and one electrovalency. (b) This extension of the theory of intramolecular ionization brings the reactivity of organic compounds into line with the well-recognized activity of inorganic ions, and makes it possible to regard all chemical action as ultimately ionic in character. (c) The resting forms of the molecules need not be identical with their ionized or reactive forms. Examples are given, however, of organic compounds which probably have a permanently ionized structure, comparable with that of metallic salts. In other cases evidence is quoted to show that this condition results from a definite process of acti vation. (d) Attention is directed to several phenomena which receive a simple explanation in the "crossed polarities of compounds which have hitherto been classed as "conjugated." (e) The properties of "multipolar ions" are referred to as furnishing a basis for a novel theory of tautomeric ions, and an explanation of their readiness to yield coordinationcompounds. University Chemical Laboratory, Cambridge. * Trans. Chem. Soc. cxxiii. p. 830 (1923). CII. On the Theory of Light-Quanta. By H. BATEMAN, Professor of Mathematics, Theoretical Physics, and Aeronautics, California Institute of Technology, Pasadena *. IN Na justly famous paper, in which he proposed his theory of the photoelectric effect, Einstein † showed that the idea of light-quanta is a plausible one when considered from the point of view of cavity radiation. For small densities of radiation of a definite frequency v, Wien's radiation formula may be regarded as applicable, and it appears that the radiation of this frequency behaves as if it consisted of discrete quanta, of energy hv, spatially independent of each other. Einstein's analysis has been recently extended by Wolfke ‡, who shows that Planck's radiation formula can be interpreted to mean that cavity radiation consists of a series of partial radiations thermodynamically independent of each other, the constituents of each order being distributed as regards frequency according to a law analogous to that of Wien. The density u of radiation whose frequency lies between v and v+dv is in fact, by Planck's law, is the density of a partial radiation of order s. By a repetition of Einstein's argument, in which use is made of the formula = for the entropy S of a partial radiation of orders and energy E. Vu, contained in a volume V, Wolfke finds that the probability W, that this energy may find itself at an arbitrary instant of time in the partial volume V' is • Communicated by the Author. Phil. Mag. S. 6. Vol. 46. No. 275. Nov. 1923. 3 R |