metal concentration in the alloy. The relation between | metallic copper is eo ipso constant. and the concentration these two quantities is the following:-Let the potential of the H ions constant by hypothesis, and we obtain of the alloy with regard to the electrolyte be during a finallyshort interval, while the current is flowing. The work to be done in order to transport the quantity of electricity 1 from the electrode to the electrolyte will be The ratio of the increase in the oxygen concentration to the decrease in the metal concentration of the metal oxygen alloy may be considered equal to the ratio of these two quantities of electricity; hence from which we find - kot c = e de (ki(t)k c)dt. . (13) kot (14) where C, is the integration constant. From the initial condition (for which = o) we see at once that the integration constant is dependent only upon that oxygen concentration which corresponds to the electrode when no current is flowing. This value of c is in our case given by the (7) equilibrium potential of the copper electrode. If we now consider that the chemical polarisation of a type of electrode such as we deal with in the copper electrode will, at small current density already, assume relatively large values, which differ strongly from the equilibrium potential of the base electrode, C, may be regarded as infinitely small with respect to the c values, and the equation may be written in the form (8) (9) Or in words: When with the anodic treatment of a metal its concentration in the metal oxygen alloy becomes small, then the system O-OH' will effect the whole transport of electricity: r therefore becomes equal 1. The electrode now transports the electric current in a way, as if it were a highly noble metal; its capability of primarily forming ions is completely suppressed, and it is limited to a secondary formation of ions. Let us further assume that the total current in the anodic treatment of a copper electrode in the presence of Hand of Cu ions of high concentration be given by the reaction 20H'+2 H2O+0. This theory, which very satisfactorily characterises the chemical polarisation on condition that all the metals become primarily active as anodes, fails in many instances. Without entering into detail I will emphasise the consideration which will help us on to the right way. We have seen that in the consideration of the anodic treatment of a metal the theory leads to quasi-secondary formation of ions, although in principle the theory starts from primary anodic formation of ions. That is just the desired result of the theory. This result is not arrived at, however, when we attempt to explain the cathodic We will also suppose that the slow process of chemical chemical polarisation by the aid of this theory. For if we compensation is defined by Cu + 2H +0 Cu + H2O. Let this process be much slower than the replenishing of the OH' ions which are consumed by the current. And finally let the experimental conditions be such that the rate of percentage rise in the Cu ion concentration at the electrode within the time t is so small that it may be neglected. If, then x be the total oxygen concentration which the current has supplied up to the moment t; * that concentration of the oxygen which has chemically been bound up to the moment t by Cu and Hions; c-x-x the oxygen concentration on the electrode surface at the moment t; ko the velocity coefficient of the reaction: Cu + 2H +0 Cu+H20. imagine that the two chief components of the electrode volume consist, in the cathodic polarisation of the Cu CuSO4 electrode, of Cu and of H atoms, then their quotient is constant, not their product, and the increase in the concentration of the one calls forth a simultaneous increase in the concentration of the other. Further-and that is the chief drawback of the theorywe cannot, with its aid, come to the above-mentioned type of two reactions, of which the one is retarded by a rise in the O concentration whilst the second increases its velocity in a regular manner. This type of two reactions, we have recognised, can alone elucidate the phenomena of passivity. We can easily comprehend that we shall arrive at such a type of reaction if we imagine some reaction to take place within the electrode volume, i.e., between its constituents. The first theory does not admit of that, because there is always equilibrium within the electrode volume; and all the compensation processes of it a variable current intensity for which the same direc- which the first theory admits are indirect reactions. In tion is, however, always maintained; and if, as stated, the oxygen concentration on the electrode surface at the moment t be order, therefore, that we may imagine a larger number of possible compensation processes, the mechanism of the electrode volume should be so designed that there should not prevail equilibrium in it while the current is flowing. The entire current must be transported by one system from the electrode to the electrolyte. We also saw that the prototype of the mechanism of chemical polarisation, the Pd Zn experiment, renders it necessary to ascribe to the zinc which enters the electrode volume a diminution of the concentration of the Pd atoms. (12) How is this to happen? This effect is difficult to conceive if we regard the electrode volume as constant, and if we make no new hypothesis as to the mutual influences of the various constituents of the electrode volume. The simplest assumption which at once enables us to describe all the experiments mentioned is this: The sum of the concentrations of all the constituents of the electrode volume is always constant, whether the current flows or not. is at any moment proportional only to the actual oxygen concentration because the concentration of the 6. The Constant-Sum Hypothesis and the Second Theory. (Reichenstein, Zeit. Elektrochem, 1913, xix., 672). We presume that the electrode volume is always fully saturated, and that any new constituent, possibly introduced by the electric current, ejects part of the old substances. For the present I am not in a position to maintain the idea of a primary activity of all the systems. In view, however, of the fact that this theory satisfactorily answers all the questions concerning the chemical polarisation, as well as those concerning the spontaneous solution of solid bodies in the presence of oxidising agents, we may for the present rest satisfied with the assumption that in aqueous solutions the transport of the electric current across the boundary electrode/electrolyte is effected only by the OH' and the Hions. When a metal electrode is treated anodically the oxygen introduced pushes the metal atoms out of the electrode volume-the same will happen on the cathode by the H2 introduced-and the chemical equilibrium prevailing before the current was flowing will be destroyed. The oxygen now begins to react with the metal atoms of the electrode volume. Let this reaction be of the form Me | O2H→ Me" + H2O, where Me and O are respectively an atom of the metal and of the oxygen of the hypothetical alloy, and H. con. cerns the electrolyte. If the concentrations of the three substances reacting with one another are x, y, z, the velocity of the compensation process can be formulated as follows: According to the hypothesis of the constant sum (16) (17) curve 17 This equation gives us the above-discussed point of inflection for the y-t curve, that is to say, also for the curve which illustrates the rise of the polarisation with time after closing the circuit. We have now arrived at some point of reference enabling us to describe the experiments of Andrejew. When we immerse a foil of gold in a solution of KCN which contains an oxidising agent like O2, H2O2, or O3 in solution, two purely chemical reactions will come into play: I. The formation of the gold-oxygen alloy. This process is missing in the electrolytic dissolution of metals because in their instance the oxygen is directly supplied by the primary discharge of OH' ions, whilst here the oxygen is supplied by a chemical reaction (= decomposition of the oxidising agent). If the oxygen were not to disappear owing to the dissolution of the gold, we should finally arrive at an equilibrium between its concentration in the electrode volume and the concentration of the oxidising agent in the aqueous KCN solution. We may suppose that in one and the same oxidising agent these two concentrations will bear a rectilinear relation. Now the formation of the gold-oxygen alloy, we shall see, does not proceed at an infinitely rapid rate, yet not so slowly that this process should become decisive for the entire process of solving the gold. The CN' ions are, moreover, not in contact now with the compact gold metal, but in contact with the gold-oxygen alloy, and process II. is taking place. II. 2Au | O (alloy) + H2O+CN' → 2Au(CN)'2 +2OH'. This process may, of course, proceed in several stages. When we increase the O concentration of the alloy by increasing the concentration of the oxidising agent in the liquid phase, the critical O concentration will finally be exceeded, and the velocity of the process II. will diminish more and more. It should here be emphasised that there are many instances in which the rate of oxidation becomes smaller as the oxygen concentration is raised; I will content myself with drawing attention to the table of these cases com(18) piled by van't Hoff ("Vorlesungen über theoretische und physikalische Chemie," No. 1, Braunschweig, 1898, 238). The maximum corresponds to the value y = When with a certain current density of the electrolysis the compensation process is so rapid that the oxygen concentration of the electrode volume does not rise above the value (which we may call the critical concentration), a 2 then there will be no reason to fear that another com. pensation process will set in, e.g., generation of oxygen (=passivity). When, however, the critical concentration is exceeded, the compensation process will necessarily be retarded, and the O concentration will increase rapidly until a new compensation process sets in. Instead of the above-mentioned equation of the first theory, which defines the accumulation of the O concentration in the electrode volume after closing the circuit : dy - ki - koy, The critical O concentration has now been exceeded, and, in accordance with what has been explained, a new chemical process (a second compensation process) must come into play. Such processes are the following: formation of H2O2 (observed by Bodländer), formation of gold peroxide. We recognise that the occurrence of oxide layers must not be regarded as the cause of the delay in the solution of gold. The peroxide films are merely a concomitant feature, which indeed indicates that the O concentration of the alloy has attained high values, but which must by gold. It is a symptom, like the rise of temperature in the no means be represented as the cause of the dissolution of diseased human body. It is interesting to note that some investigators regard as cause of the passivity that in which other investigators see the consequence of the cause of passivity. Thus, process I. supplies oxygen to the alloy metaloxygen, and temporarily increases its concentration, whilst process II. tends to diminish this concentration. In the stationary condition, when the reactions proceed at a constant rate, the concentration of the oxygen in the alloy is dependent upon the two processes I. and II.; that is to say, the O concentration may be increased, not only by increasing the rate of process I., but also by diminishing where i is the current intensity of the continuous current, the rate of process II. Ceteris paribus, therefore, the "critical O concentration" (in the process of dissolving gold in diluted KCN) may be exceeded with the aid of (19) oxidation potentials of the oxidising agents, which are so low that they would be ineffective at higher CN' ion concentration. This is in agreement with the experiments of Andrejew, according to which the characteristic film of (20) gold peroxide appears in solutions of KCN ranging from O'01 to 0.05 per cent, already when air acts as oxidising agent, which demonstrates that process I. is absolutely will have the following spatial distribution: Electrolyte necessary. We may draw a comparison as to the spon-hypothetical alloy thin layer of the noble metal/compact taneous solution of passive and of active metals in the base metal. The noble metal may then separate from the presence of oxidising agents from this point of view. electrode. For this end, we know from galvanoplastic What could we predict as regards the spontaneous solution processes, a trace of impurity between the two metals will of copper and nickel in H2SO4 in the presence of oxidising be sufficient. agents? We know that the latter of these two metals belongs to the so-called passive metals, i.e., that the process of solution is slow. We should then expect, from what has been explained, that the critical O concentration of the nickel-oxygen alloy, and hence the retardation in the rate of solution of nickel, will, ceteris paribus, always occur at such oxidation potentials of the oxidising agent at which this will not be attainable in the case of copper; and it is altogether doubtful whether the v maximum will experimentally be realised in the dissolution of copper. Of especial interest from the standpoint of the new theory are the exposition of the cases of the palladiumzinc cathode and the behaviour of the alloys mentioned on the anode. When a palladium electrode, to the electrolyte of which a zinc salt has been added, is treated cathodically, the current is carried by the H ions. Owing to the prevailing conditions of equilibrium, some zinc will enter the electrode volume and will push out some palladium, already before the circuit is established. When the current flows, the zinc acts as ballast, because it depresses the Pd concentration according to the hypothesis of the constant sum, and hence retards the rate of the compensation proWhen the concentration of the zinc in the electrode volume is y and the sum-constant is a, the maximum of the curve, which expresses the rate of the compensation process (Pd+H reaction) as a function of the H concentration, will correspond, not to the H value,, but to cess. 2 2 a-y. Now, hydrogen accumulates more and more in the electrode volume, and the polarisation rises. It should be noted that the whole conception forces us to assume a chemical reaction between the H atoms and those Pd atoms which fill up the electrode volume-an assumption which was impossible according to the first theory of there being always equilibrium in the electrode volume. Thus quite generally any foreign substance entering into the electrode volume which will not take part in the compensation process may be regarded as ballast. A special interest attaches, from this point of view, to the behaviour of the real alloys forming on the anode. When, in the case of a purely active metal, the maximum of the curve O concentration/velocity cannot be attained, that maximum can, in the case of an alloy as described, be realised, owing to the ballast, at relatively low current density, and the curve will show an inflection. An alloyed anode, therefore, displays formally the properties of a passive metal. We may imagine the whole anodic process on the following lines. As regards the space distribution the electrode, at the beginning of the anodic treat ment, has the appearance: Electrolyte hypothetical alloy real alloy/compact base metal. The anodic treatment carries oxygen into the hypothetical alloy, where it partly displaces both the metals, and the chemical reaction between the oxygen and the base metal sets in. (As a matter of fact the two metals enter into competition with regard to the oxygen; practically, however, the oxygen reacts with the less noble metal only.) The base metal, which has disappeared owing to the chemical reaction, is being replaced in the hypothetical alloy by both the metals of the real alloy. In which ratio? That depends upon the rate of diffusion of the base metal, or upon the rate of the alloy formation of the two metals. If this rate be small, then the concentration of the base metals in the hypothetical alloy (when the current is flowing) will also be small-the velocity of the compensation process will decrease, the polarisation will rise, and finally will follow the decomposition of the real alloy; the electrode then In order to keep the rate of alloy formation alluded to small, we must, as has been pointed out already, exercise the greatest care as to the purity of the two metals in building up the real alloy. Certain foreign substances accelerate the alloy formation; nitric acid in particular should here be mentioned; it plays the part of the soldering liquid in soldering two metals. THE SCIENTIFIC WEEK. THE REDDENING OF LEAVES. Prof. Gaston Bonnier has presented to the Academy of Sciences a new study of M. Raoul Combes, concerning the crystallisable substances that are found in leaves. One of these substances of a yellow colour is found in green leaves; it is this substance which produce the reddening of leaves in autumn. M. Combes, who had obtained outside the organism the synthesis of the red substance, by starting from the yellow substance by reduction, has just obtained the inverse, that is to say, the synthetic production of the yellow substance, by starting from the red substance, by oxidation. Thus the general question, the answer to which has been so long sought after in vegetable physiology, has at last been solved. Besides its general importance in the life of plants it is easily understood that this solution may be susceptible of practicable applications, notably in what concerns the different varieties of vines. THE ORIGIN OF PETROLEUM. Learned men are not agreed as to the origin of petroleum. Two schools exist; one of which attributes the formation of petroleum to the igneous action of carbides on water; the other school considers that the petroleums proceed from a distillation of vegetable origin. Prof. Armand Gautier, following on the recent studies of MM. Aimé Pictet and Maurice Bouvier, which he has made known to the Academy of Sciences, gives the preference to the second hypothesis of a vegetable formation. The two learned Geneva gentlemen have distilled ordinary coal at a pretty low temperature, somewhere about 450° in vacuum. They have thus obtained a special kind of tar vacuum, tar without any phenol or any aromatic hydro-carbides. Washed with alkali and sulphuric acid this vacuum tar treated by sodium gives a powdery product having a discharge of hydrogen. This body dissolved in water gives birth to semi-aromatic alcohols, derived from camphor and from hydrocarbides, having the formula C10H20 or C12H22, which have the same characteristics as the Canadian petroleums, the same point of fusion, same smell, &c. These are quite new experiments and of the greatest interest. THE SPRING FINDERS. At the time of the famous experiments undertaken at the second Congress of experimental psychology by the spring finders and adepts of the divining rod, M. Armand Viré, Professor at the Natural History Museum, was exceedingly incredulous. However, he became converted, and indeed became a spring finder himself, and, as it were, in spite of himself. And it must be admitted that his conversion caused some astonishment. During the summer months he invited several spring finders to the grottoes of Padriac, with which he is specially well acquainted and of which the plans have not been published. There were several negative experiments. But two or three spring finders were remarkably successful. On the surface of the soil they followed several subterranean water springs and CHEMICAL NEWS, Rotatory Power on Chemical Constitution Jan. 9, 1914 arrived at the points where they left the earth and appeared on the surface. These points were previously unknown by the spring finders. M. Edmond Perrier, when presenting these experiments to the Academy of Sciences, added that neither he nor M. Armand Viré take upon themselves the responsibility of these observations. They offer no explanations, but simply give the mere facts. THE MAP OF the World to a Scale of ONE MILLIONTH. M. Charles Lallemand, Director of the Service and Organisation of the General Levelling of France, has given some information concerning the last conference of the international map of a millionth scale. At the first conference, which was held at London in 1909, only seventeen countries were represented. At the present time thirty-one states have joined the convention. Lallemand also announced that the Prince of Monaco was M. taking upon himself the expense of several oceanic pages. MICROBIAN LIFE. Microbian cultures are often very difficult. M. Adrien Lucet, Member of the Academy of Medicine, has just shown that a regular agitation or shaking of the liquid mediums used in bacteriology, contrary to the opinion that is generally admitted, acts favourably on these infinitely little creatures. By making these bouillons undergo a slow and continuous movement he has, in fact, been able to obtain cultures as many as eight times more abundant of the microbes of cholera, typhus fever, carbuncle, diphtheria, the glanders, dysentery, and even of lock-jaw, the microbe of which can only be cultivated without the penetration of any air. It was M. Chauveau who communicated M. Lucet's study to the Academy. PROCEEDINGS OF SOCIETIES. CHEMICAL SOCIETY. MESSRS. R. E. Slade and S. C. Sastry were formally admitted Fellows of the Society. Certificates were read for the first time in favour of Messrs. Albert Frederick Calvert Royston, Eton Avenue, N.W.; Behari Lal Das, 107/2/1, Manaharpukur Road, Kalighah, Calcutta; Eric Russell Harrap, Maisemore, Ebury Road, Rickmansworth, Herts; Oswald Ryle Horwood, M.A., M.R.C.S., L.R.C.P., Tunstall Rectory, Suffolk; Dan Ivor James, M.A., B Sc., Frondeg, Llandilo, Carmarthenshire; Alexander Williamson McLaren, 3, Hayfield Terrace, Langside, Glasgow; Harold Edwin Temple, 239, Cashel Street, Christchurch, New Zealand; Robert James Wright, M.A., care of R. Burnett, Esq., 336, Pollokshaws Road, Glasgow. A Certificate has been authorised by the Council for presentation to ballot under By-law I.(3) in favour of Mr. Bertie Mandel Welsh, 80, Hunter Street, Sydney, N.S.W. The PRESIDENT announced:-1. That the bust of the Rt. Hon. Sir Henry Roscoe which was exhibited at the meeting had been presented to the Society by the friends and former students of Sir Henry Roscoe. 2. That by request of the Council Messrs. Vieweg and Sohn have offered to sell the first volume of the "Literatur Register," by R. Stelzner, to Fellows of the Chemical Society at the reduced price of £3 10s. (original price £4 4s.), provided that not less than twenty copies are sold to Fellows of the Society. Those Fellows who desire to obtain a copy of the "Literatur Register these terms are requested to send in their names to the Honorary Secretaries. on 19 The PRESIDENT referred to the meeting of the International Association of Chemical Societies which had been held in Brussels during September, 1913, and drew the attention of Fellows to the abbreviated report of the meeting which the Council have ordered to be printed in the Proceedings. * Of the following papers, those marked were read :*286. Investigations on the Dependence of Rotatory Power on Chemical Constitution. Part V. The Simpler Esters of the Carbinols, CH3 CH(OH) R." BY ROBERT Howson PICKARD and JOSEPH KENYON. aliphatic acids have been prepared and examined polariThe following homologous series of esters of normal metrically under various conditions:-(i.) The esters of d-methylethylcarbinol; (ii) the esters of d-methyl-npropylcarbinol; (iii.) the esters of d-methyl-n-butylcarbinol of d-methyl-n-hexylcarbinol; (vi.) the esters of d-methyl-n(iv.) the esters of d-methyl-n-amylcarbinol; (v.) the esters nonylcarbinol [these six series range from the acetates to the stearates]; (vii.) the acetates and (viii.) the n-dodecoates of the d-carbinols [methylethylcarbinol to methyl-nundecylcarbinol j. Many of these esters (all of very simple constitution) metrically at temperatures above 150° or at various conexhibit anomalous dispersion when examined polaricentrations in solvents such as pyridine, benzene, or carbon disulphide. *287. "Co ordination of Rotatory Powers for different Wave-lengths, Temperatures, and Solutions." (Preliminary Note). By ROBERT HOWSON PICKARD and JOSEPH KENYON. The authors have already described the synthesis of thirty optically active carbinols of the general formula RICH(OH) R, and some seventy esters derived from some of them. These compounds have been examined polarimetrically for three wave lengths (a) in the homogeneous state at different temperatures, and (b) in several solvents at various concentrations. It was found (see preceding paper) that many of these under certain conditions exhibited anomalous optically rotatory dispersion. The attention of the present authors was therefore directed to a paper by H. E. Armstrong and E. E. Walker entitled "The Causes of Variation in the Optical Rotatory Power of Organic Compounds and of Anomalous Rotatory Dispersive Power" (Proc. Roy. Soc., 1913, A, lxxxviii., 388). In this paper it is suggested that anomalous dispersion is caused by the presence of two substances (in the actual cases considered of two isodynamic forms) having rotatory powers of opposite sign and different dispersive powers. In the case of esters of such simple constitution as those described in the preceding paper, the suggested explanation of the anomalous dispersion seems feasible, however, only on the assumption of a change in the association of the esters, not only in the homogeneous state on increase of temperature, but also on solution in various solvents. A "characteristic diagram" for d sec.-octyl acetate was therefore coustructed according to the method of Armstrong and Walker (loc. cit.). A reference line with slope of unity was drawn, and on it were plotted the various numbers representing the specific rotatory powers for mercury-green light. The numbers representing the specific rotatory powers for sodium-yellow and mercuryviolet lights were then plotted on the ordinates passing through the points previously located on the reference line. The various points for the latter two lights were found to lie on two straight lines, and the diagram was similar in character to those for the substances of previously known anomalous dispersive power as drawn and described by Armstrong and Walker. It is now found, however, that this same "characteristic diagram" can be used to co-ordinate the results of all the determinations of rotatory power of one of the two optically active forms of the hundred synthetical com pounds previously described by the present authors; thus the numerical results (varying from +50 to -25) of the determinations of rotatory power for the three lights in different solvents at all concentrations and in the homogeneous state at different temperatures, not only of one compound, but of many (all of which are of very simple, but closely related constitution), have been plotted on one diagram. In this the various values lie on three straight lines, which intersect above the zero-line, and not all at one point. The dispersions then appear in the diagram as some function varying with the magnitude of the rotatory powers, thus co-ordinating the small, but very definite, differences in the dispersions of the homologous compounds which have been observed experimentally. 1882, xv., 670). Exposure to radium bromide also in creased the activity. This is in accordance with the work of Kailan (Monatsh., 1912, xxxiii., 1329). DISCUSSION. Dr. SENTER Suggested that the results would probably come under the heading of over-voltage phenomena, the best-known example of which was the retarded action between pure zinc and sulphuric acid. Over-voltage at the boundary between solid and liquid appeared to be connected with surface tension, and it might be assumed that the active substance, whatever it might be, modified the surface tension. With reference to the author's proof of the presence of It cannot yet be said whether this co-ordination is due liberation of iodine from iodides were rather untrusthydrogen peroxide in tap water, tests depending on the either (i.) to the comparison of a large number of com-worthy, as in certain circumstances dissolved oxygen gave pounds which have very similar dispersive powers, or (ii.) to the closely related constitutions of the substances. This method of plotting appears to afford (in some cases at least) a means by which a derivative differing in sign from that of the optically active parent substance can be properly designated "d" or "," and also a means of determining whether a change in configuration has taken place in the formation of a derivative. DISCUSSION. Sir W. RAMSAY suggested that the cause of rotation was ultimately in the direction of the circular motion of electrons within the molecule. A dextrorotatory substance, say, might have hydroxy in the molecule, capable of repelling an electron, and causing it to rotate dextrorotatorily. On the other hand, if the hydroxyl be replaced by bromine, the bromine atom might have the property of attracting an electron and of reversing the direction of its rotation. This suggestion was made, not with any conviction of its applicability, but merely for consideration as to whether it was possible to obtain any clue to the fundamental reason of rotation. *288. "The Interaction of Sodium Amalgam and Water." By HERBERT BRERETON BAKER and LESLIE HENRY PARker. Water distilled under special conditions has a visibly slower rate of action on sodium amalgam than ordinary distilled water. An apparatus was constructed to measure accurately the hydrogen evolved, and various samples of water were tested. The least active specimens of water were obtained by distillation from copper or platinum apparatus, especially on superheating the steam before the latter was condensed. The rate of action was shown to be no function of the conductivity of the water used, but was found to depend largely on the pressure at which the reaction was conducted, increase of pressure causing the rate of action to diminish, and vice versa. Various explanations of this phenomenon were put forward and tested, but the only feasible one was the assumption of the presence of some impurity in minute quantity, which was volatile under the conditions of ordinary distillation, but was destroyed on heating to redness. Experimental evidence was adduced which seemed to show that the only impurity which satisfied all the conditions was hydrogen peroxide, and the widely differing activities of various samples of water on sodium amalgam is ascribed to the presence of varying quantities of hydrogen peroxide. *289. "The Action of variously treated Waters Sodium Amalgam." By LESLIE HENRY PARKER. on Further evidence was adduced in support of the explanation of the varying activity of different samples of water on sodium amalgam put forward by Baker and Parker (preceding paper). Various metals were sealed up with a quantity of the inactive water for definite periods. Metals such as copper, mercury. &c., did not have much effect, whilst aluminium increased the activity of the water on sodium amalgam. This was shown to be in harmony with Traube's work on the wet oxidation of metals (Ber., | the reaction in question. The titanium dioxide test was trustworthy, and was extremely sensitive, as it was capable of detecting I part of peroxide in 50 million parts of water (compare Senter, Trans. Faraday Soc., 1906, ii., 142), and the fact that it was not given by the water in question appeared to render further investigation desirable The effect of added hydrogen peroxide might be connected with the fact that this compound was readily decomposed at a mercury surface (Bredig); the evolution of oxygen would presumably disturb the very unstable equilibrium characteristic of over-voltage. Dr. R. E. SLADE agreed that it was to be expected that hydrogen peroxide would disturb the over-voltage at the surface of the amalgam in the manner suggested by Dr. Senter. He believed, however, that a very important factor was the existence of dust particles in the water, and quoted the work of G. N. Lewis on the potential of sodium amalgams in support of this. The experiment which Mr. Parker has just shown rather pointed to this theory, for in the tube of pure water the bubbles of hydrogen came off at a few points which moved about on the surface of the amalgam. Perhaps the superheating of the steam was an efficient way of removing particles of dust by destroying them or by causing them to adhere to the hot tube. been studied in connexion with the experiments described, Dr. KEANE asked whether the influence of light had as this might have considerable effect both in regard to the production and decomposition of hydrogen peroxide. In reply to Dr. Senter, Mr. PARKER said that no comparative experiments had been made as to the relative values of the tests for hydrogen peroxide: further experiments were in progress with the view of obtaining some light on the mechanism of the reaction. In reply to Dr. Keane, he also stated that no experiments had been tried on the influence of light on the reaction. GEORGE FRANCIS MORRELL and PETER BURGEN. *290. "The Polymerisation of Cyanamide." By The polymerisation of cyanamide, under various eondi. tions, both in the solid state and in solution, and also under the influence of catalysts, has been quantitatively studied. With the pure substance itself only about 10 per cent was found to have changed in six months, and in aqueous solution, even at elevated temperatures, the reaction proceeds comparatively slowly, many hours' heating at 100° being required to complete it. In absolute alcoholic solution the reaction-velocity is much further reduced. In all these cases the velocity-constant was not that of a bimolecular reaction, but equal amounts were found to be changed in equal intervals, except at great dilution, in which case a logarithmic curve was obtained. An ionic explanation may be advanced to explain these facts, the ions present in very small and, at first, practically constant concentration alone taking part in the change. The influence of acids and bases, such as sulphuric acid, ammonia, and sodium hydroxide, as catalysts produces an extremely marked acceleration, very small quantities reducing the period of half-change in aqueous solution at 100° from about twelve hours to as many |