ii. 45 General and Physical Chemistry. The Secondary Spectrum of Hydrogen. A. C. MENZJES (Nature 1922 110 876).-Silberstein’s solution of the three-substance problem applied by him to the case of neutral helium has been modified so as to apply to hydrogen and the value of N has been corrected so as to take account of the fact that with two electrons instead of one the correction to the mass of the electron for the finite mass of the nucleus is no longer the same. Frequencies have been calculated from the formula v = N ~ (1 /n12-f 1 /n22- l/m12- 1 /m22) where N =N,( 1 +m/M). Forty-seven lines in the secondary spectrum of hydrogen were found to agree with the calculated values within an absolute error of one unit of frequency. The frequencies are regarded as a kind of “summation tone,” being the sums of a Balmer or a Paschen frequency and a frequency in the ultra-red.In several cases a physical similarity of behaviour was cummon to “ series ” of the lines grouped according to values of m and n. Spectrum of Active Nitrogen as Affected by Admixture of the Inert Gases with a Note on the Origin of the Cyanogen Spectrum. LORD BAYLEIGH (Proc. Roy. Soc. 1923 [A] 102 453459).-The afterglow accompanying the reversion of active nitrogen to ordinary nitrogen was shown by E’owler and Strutt (A. 1911 ii 678) to be composed of a selection of the first positive bands of nitrogen. The addition of inert gases to the nitrogen produces considerable changes in the intensities of these bands. The maxima of the three groups of bands green yellow and red are shifted towards the red.The addition of helium neon or argon makes the red group as a whole; more intense at the expense of the others. The movement of the bands is proportional to the concentration of the inert gases present. Evidence is adduced for the belief that the red aurora line X 6320 is due to the presence of nitrogen and helium in the upper atmosphere. It is not yet determined whether the effects on the after-glow spectrum are produced in the electric discharge or in the process of the after-glow. The influence of carbon in developing the cyanogen bands is discussed. W. E. G. The Carbon Arc Spectrum in the Extreme Ultra-violet. I?. SmmoN (Proc. Roy. SOL 1923 [A] 102 484496).-Measure- ments are made of the ultra-violet lines of the carbon arc spectrum by means of a vacuum grating spectrograph.Colour sensitised Schumann plates were used both for the extreme ultra-violet as well as for the ordinary violet region. A table is given of the wave- lengths and intensities of the lines observed and comparison made with the values of other workers. Some of the lines photographed in the Lyman region have not been mentioned by any previous A. A. E. VOL. CXXIV. ii. 3ii. 46 ABSTRACTS OF CHEMICAL PAPERS. worker and some of the lines in the higher orders of spectrum have been shown to consist of close groups of lines. The line 599.0 attributed by Lyman to helium is due to carbon present as an impurity in the helium. The lines in the Lyman region a t 1194,945 858 687 651 640 599 and 595 have not been previously observed in the arc spectrum of carbon but these lines correspond with prominent lines in the " hot spark " spectra of Millikan.Groups a t 1657 1329 1260 1194 1036 and 651 have been resolved into a number of lines. W. E. G. The Spectrum of Neutral Helium. LUDWIK SILBERSTEIN (Nature 1923 111 46-47).-1n reply to Raman's criticisms (A. 1922 ii 803) it is shown mathematically that the probability that the forty-five coincidences previously described are fortuitous is less than 1.7~10-13; it is further claimed that the use of the value 109723 for the Rydberg constant in the formula applied to the case of the neutral helium atom (with two electrons) is justifiable. It is now found that the whole diffuse series of singlets 1P-mD is represented by the formula v=4N(18 .2n/20.4)E-N(9.n/lO .2) two final and one initial quantum numbers being fixed. The possibility of reducing 4N to N based on the fact that all numbers are even is mentioned but not discussed ; physically interpreted it would mean that the helium nucleus attracts each of its electrons with only one-half of its total charge as if its lines of force formed two bundles each entirely engaged with one of the two trabanh. A. A. E. Relative Visibility of Spectra when an Electric Discharge is Passed Through the Vapours of Alkali Amalgams. F. H. NEWMAN (Phil. Mag. 1923 [vi] 45 181-189).-The relative intensities of the spectra of the alkali metals and mercury have been examined by placing alkali amalgams in an electric discharge tube. ' For sodium and mercury the spectrum of the former was prominent a t all temperatures but below 100" the mercury spectrum was the brighter.At 200° the mercury lines are almost entirely masked. At this temperature a very brilliant yellow radiation is emitted which forms a very convenient source of the sodium D-lines. The ionising potentials of sodium (5.13 volts) and mercury (10.2 volts) and the partial pressures of their vapours are the two main factors operative in determining the intensity of the spectra. At low vapour pressures the atoms of both sodium and mercury will be so far apart that the electrons attain the necessary energy to ionise the mercury atoms before inelastic collision occurs. Under these conditions approximately equal numbers of mercury and sodium atoms will be ionised. As the temperature rises the density of the vapour increases and the electrons attain less frequently the speeds corresponding with 10.2 volts so that although the number of mercury atoms will be far more numerous that those of sodium the actual number of the former ionised will be very small.Thus a t 200" the mercury lines are suppressed. Except in the case of potassium where theGICNBRAL AND PHYSICAL CHEMISTRY. ii. 47 lines are faint a t all temperatures the other alkali amalgams show similar phenomena. W. E. G. The Excitation of the Mercury Spectrum. The Influence of Helium. GEORGES D~JARDIN (Compt. rend. 1922 175 952-955).-An examination of the radiations emitted by mercury vapour traversed by electrons of different velocities showed that when the speed of the electrons is slightly greater than the critical ionisation velocity a large number of the arc spectrum lines are obtained these being due to collisions between electrons and mercury atoms.When the conditions are such that the electron charge is neutralised a luminous region appears near the electrodes the spectrum of which also exhibits the characteristics of that of the mercury arc. The author concludes that the emission spectrum appears to depend on the ionisation of the atom and notes the conditions under which certain of the lines become visible. In the presence of helium and with a potential greater than 20.4 volta considerable modifications of the mercury spectra take place. Details of the variations in intensity and the appearance of new lines are given the action of helium being apparently selective on certain lines as in the case of cadmium (cf.Collie and Watson A. 1918 ii 383). H. J. E. Spectroanalytical Investigation of a New Element in the Terbium Group and the Arc Spectrum of Terbium. JOSEF MARIA EDER (Sitzungsber. Akad. Wiss. Wien Math.-naturwiss. KZccss. 1922 [iia] 131 199-298; cf. ibid. 1920 129 422).- The determination of the spectra of a number of terbium dysprosium and gadolinium fractions has led to the discovery of a new element giving a well-defined and characteristic line spectrum. This element which was present in a fraction between terbium and dysprosium is named Welsium. The lines of the new element are weakened in the spectra of the fractions rich in terbium and are practically absent in the spectra of the dysprosium fractions.In agreement with the observations of Eberhard no evidence has been found for a new element in the fractions between terbium and gadolinium. Many thousands of lines of the terbium spectrum have been measured between A7257 and 2400 8. and compared with the measurements of Urbain Eberhard and Auer. W. E. G. The Line Spectrum of Vanadium in Fused Salts. A. DE GRAMONT (Compt. rend. 1922 175 1129-1133; cf. A. 1921 ii 73).-A solution of vanadium pentoxide in fused sodium carbon- ate was used in order to ascertain the limiting quantity of vanadium capable of spectroscopic detection. The results are shown in two tables which give for visual and photographic observation re- spectively the smallest quantity corresponding with each line in the spectrum.Non-conducting materials which contain vanad- ium give good spark spectra which are in accordance with the author's observations. H. J. E. 3-2ii. 48 ABSTRACTS OF CHEMICAL PAPERS. Absorption of Light by Chlorine. H. VON HALBAN and K. SIEDENTOPF (2. physikal. Chem. 1922 103 71-90).-The absorption of light by chlorine between 254pp and 643pp has been determined for nineteen wave-lengths using lines of mercury zinc and cadmium and between 383 pp and 451 p using a Nitra- lamp for twelve wave-lengths. The method of observation was the previously described photoelectric two-cell arrangement (A. 1922 ii 332). The results show that within the pressure range investigated (760 mm. and downwards) Beer's law is true. The absorption curve shows a maximum in the ultra-violet at about 334pp and a minimum in the yellow.From yellow to red the absorption again increases and reaches a maximum which probably lies between 614 pp and 643 pp. No difference could be observed between the absorption of dry and moist chlorine. The Ultra-violet Absorption Spectra of Toluene and the Xylenes. F. W. KLINGSTEDT (Compt. rend. 1922 175 1065- 1067).-The ultra-violet absorption of hexane solutions of toluene and of the xylenes was measured and details of the position and form of the bands are given together with a diagrammatic repre- sentation. Toluene exhibits sixteen bands as compared with eight for benzene in the same region; the intensity of absorption is a little greater than that of benzene. The bands are divided into four groups in each of which intensity decreases with wave- length.Three large bands are observed with o-xylene and five with m-xylene whilst in the case of p-xylene the absorption coefficient is about four times as great and the spectrum contains twelve bands arranged in three groups. Colour and Chemical Constitution. XV. A Systematic Study of Fluorescein and Resorcin-benzein. JAMES Mom (Trans. Roy. Xoc. S. Africa 1922 10 159-164; cf. A. 1922 ii 333).-The quantitative study of the colour of compounds of the phenolphthalein type has been continued. The results pre- viously recorded for the bromophenolphthaleins (A. 1921 ii 365) are restated in another form. The different positions open to substitution in the phenolphthalein molecule are lettered as shown. The colour modification due to the introduction of bromine into any particular position can be expressed by a factor the '' dicyclic colour factor,'' by which the characteristic H 0 / Y l ./\OH wave-length of phenolphthalein must be 31 I oHI I multiplied to give that of the derivative.These factors for the respective positions are a and d 1.018; b 1-002; c f and g 1.0145; e and h 1-0270; i and I 1.0235; j andk 1.0125. If two or more positions are substituted all the corresponding factom must be used in calculating the colour of the derivative. The colour factors for chlorine and iodine are respectively about 1/1500 smaller and greater than those for bromine. Similar factors have been worked out for various substituenb J. F. S. H. J. E. f \/\ 6 /\h/g d()co2H $ 1 "\/" bGENERAL AND PHYSICAL CHEMISTRY. ii.49 in fluorescein X=493.5. For bromine the factors for the different positions are a and d 1.0162; b 1.0000; c 1.0091 ; f and k 1.0122; g and j 1.0157. The factors for nitro-derivatives are very similar to those for bromo-derivatives the methyl factors are smaller and the methoxy-factors smaller still the largest being 1.006. The hydroxy-factors are smallest of all and in sodium hydroxide solution the hydroxy-derivatives have their absorption bands in the same position as fluorescein itself. Gallein and hydrQxyquinolphthalein are exceptions to this rule. Resorcin-benzein which is fluorescein without the carboxyl group is practically identical with fluorescein in alkaline solution ; the absorption band is at h492. A few derivatives which have been examined show that the colour factors for the f g j and k positions are the same as those of phenolphthalein.In the phenyl ring the effect diminishes in the order d c b but is still positive for b although in fluorescein it appears to be nil for this position. E. H. R. Ultra-violet Absorption Spectra of Alkaloids of the iso- Quinoline Group. Papaverine and its Hydrochloride. PIERRE STEINER (Compt. rend. 1922 175 1146-1149).-A com- parison of the absorption spectra of papaverine with those of iso- quinoline and veratrole the substances of which the alkaloid is constituted shows that it is similar to that of isoquinoline but simpler. The simplification appears to be due to the veratrole group which also determines a displacement towards the red and an increase in absorption.The spectrum of the alkaloid is similar in ethereal and in alcoholic solutions; in the latter a shifting of four bands towards the red is observed. The combination with hydrochloric acid causes a fusion of three absorption bands exhibited by the alkaloid together with a general displacement towards red and increase in absorption. The spectrographic method applied to the detection of papaverine permits of the recognition of 0.03 mg. in 2 C.C. of solution. Kinetics of Photochemical Reactions. RUDOLF WEG- SCHEIDER (2. physikal. Chem. 1922 103 273-307).-A theoretical paper in which the laws of photochemical kinetics are developed on the basis of the assumptions made for ordinary chemical reactions. Van’t Hoff’s law of the proportionality between the amount of chemical change and the amount of light absorbed holds when the conversion of only one molecular species into the reactive condition by the light determines the velocity of the reaction; that is the spontaneous return of the active molecules into the inactive form may be neglected and all other partial reactions of the total change take place instantaneously.It is to be expected that the law will not hold for the total reaction if‘the photosensitive molecular species is in excess of the other molecular species participating in the reaction; t’his is particularly the case in changes which have proceeded nearly to completion. Einstein’s equivalent law determines for energy storing reactions the velocity constant of van’t Hoff’s formula or gives a t least the upper limit H.J. E.ii. 50 ABSTRACTS OF CHEMICAL PAPERS. for the total reaction. An upper limit of the velocity only can be obtained by thermodynamical methods for energy storing reactions. Velocity equations have been integrated on the assumption of van’t Hoff’s law and a homogeneous parallel beam of light for a system which is continuously agitated and in which the light is absorbed either by a molecular species which is being destroyed by the reaction or by a sensitiser which may have a constant concentration or may be produced by the reaction. A number of other possibilities and conditions in the reactions are also theoretically considered. It is also shown that when in a given reaction two photosensitive molecular species are present it is not likely that the velocity of reaction will be proportional to the product of the two quantities of absorbed light but rather is a proportionality between the velocity and the second or higher power of the light intensity to be expected if a slower subsequent reaction requires the participation of several activated molecules.J. F. S . Photochemical Decomposition of Hydrogen Peroxide. CHR. WINTHER (Canske Vid. Selsk. illat.-phys. Medd. 1920 2 3-18 ; from Chem. Zentr. 1922 iii 981).-Hydrogen peroxide in the presence of potassium ferrocyanide in dilute solution is decomposed on illumination with ultra-violet light. The reaction proceeds a t first slowly then more rapidly and finally slows down. This decomposition of hydrogen peroxide is attributed to the formation of a catalyst by the influence of light on potassium ferrocyanide solution.The production of the catalyst increases with the time of illumination and reaches a maximum which depends on the amount of energy of the ultra-violet radiation and also to some extent on the concentration of hydrogen peroxide. G. W. R. Spacial Progression of Photochemical Reactions in Jellies. A. BENRATH and K. SCHAFFGANZ (2. physikal. Chem. 1922 103 139-154).-The photochemical actions between ferric chloride and tartaric acid silver bromide and chlorine water have been examined when the reacting substances were uniformly distributed through various jellies such as silicic acid starch egg-albumin gelatin and various animal membranes. It is shown that the progression of photochemical reactions in jellies proceeds according to Lambert’s law if the jelly absorbs the active rays but if the jelly does not absorb these rays then the progression of the reaction is proportional to the time.From the experiments on the decom- position of chlorine water i t is shown that the photosensitive com- ponent is the hypochlorite-ion. Organic jellies are shown to possess a considerable transparency for ultra-violet light whilst animal membranes allow only a very little of the ultra-violet light to pass through them. J. F. S. Photocatalysis. 111. The Photosynthesis of Naturally Occurring Nitrogen Compounds from Carbon Dioxide and Ammonia. EDWARD CHARLES CYRIL BALY ISIDOR MORRIS HEILBRON and HAROLD JACOB STERN (T. 1923 123 185-197).GENERAL AND PHYSICAL CHEMISTRY.ii. 51 Spectrophotoelectrical Sensitivity of Argentite (Ags). W. W. COBLENTZ (Bull. Bur. Standards 1922 18 265-280; Xci. Paper No. 446).-A study of the effect of crystal structure on the photoelectrical sensitivity of silver sulphide. A comparison is made between the results for acanthite (A. 1920 ii 212) and those now obtained for argentite. Appreciable changes in the electrical resistance of argentite and acanthite occur when these substances are exposed to wave-lengths of light extending from 0.3-2 p maxima being obtained at 1-35 p and 0-41 p. Argentite reacts slightly to radiation of wave-lengths 0-5 to 1.1 p whereas acanthite shows a strong photoelectrical sensitivity in this region. The maximum at 1-35 p is symmetrical in the former and unsym- metrical in the latter case.The maximum shifts to short wave- lengths a t low temperatures. The photoelectrical reaction of argentite differs from that of acanthite in being free from an induced photonegative polarisation. On increasing the intensity of the radiation in both cases a more rapid change is produced in the long wave-lengths than in the short wave-lengths and the maximum photoelectrical sensitivity is shifted toward the long wave-lengths. Mechanical working of the crystals of acanthite and argentite lowers the photoelectrical sensitivity. The worked specimens of these minerals give practically identical infra-red maxima a t low temperatures the effect of temperature being less than in the naturally occurring crystals. Thus apart from the effect of crystal structure silver sulphide has a characteristic photoelectrical response spectrum.Crystal structure has however a marked effect upon photoelectrical sensitivity. W. E. G. The Excitation of Characteristic X-Rays from Light Ele- ments. J. C. MCLENNAN and (Miss) M. L. CLARK (Proc. Roy. Xoc. 1923 [ A ] 102 389-410).-Following the method of Hughes (A. 1922 ii 184) the critical absorption wave-lengths of the K- and L-series were determined for the elements boron glucinum and lithium and the critical absorption wave-lengths of the L-series for carbon. The following values were obtained carbon L-series X=166-7 A. ; boron K-series X=83.6 8. L-series A= 292.2 8. ; glucinum K-series X=llS.2 8. L-series X=428.1 8. ; lithium K-series X=290-8 pi. L-series X=1019.0 8. The result for the K-series for boron is in good agreement with that obtained by Hughes but considerable difference occurs between the re- spective values for the L-series.The critical absorption K-wave- lengths for lithium represent the first two members of a series with a frequency formula given by v=N( 1 - 1 /m2) beginning approx- imately a t 387.7 8. and extending to 290.8 8. The model of the atom put forward by Bohr makes no provision for this series. The wave-lengths for the K-series for glucinum would extend from about X=157.6 8. to X=118.2 A. and the configurations 4(3)4(21. would appear to provide a spectral series with limits approximating to the above wave-lengths. For the elements from potassium to glucinum the square rootsii. 52 ABSTRACTS OF CHEMICAL PAPERS.of the critical potentials for the K-series are very closely propor- tional to the respective atomic numbers but lithium helium and hydrogen show a departure from the linear relation. The results however support the view that the Lyman ultra-violet series for hydrogen is the K X-ray series of this element and also that the convergence wave-length of the K-series for helium is approximately X=485.5 A. The Moseley law which is known to apply for the L-series of the heavier elements ceases to apply for elements lighter than argon but a linear relationship obtained between the atomic numbers of the lighter elements and the exciting voltages of their L-series merges into the ordinary Moseley law when the element argon is reached. W. E. G. Corpuscular Spectra and the Photoelectric Effect.MAURICE DE BROGLIE and LOUIS DE BROGLIE (Compt. rend. 1922 175 1139-1141).-Whiddington has stated (Phil. Mag. 1922 [vi] 43 1126) that corpuscular rays were not observed when h(v-v,) is less than hv v being the frequency of the exciting radiation and V that of the critical discontinuity c. The authors point out that under these conditions the rays are much less easy to obtain. Further Whiddington’s generalisation that the K-rays of tungsten do not excite corpuscular rays in the case of elements of atomic number greater than 50 does not appear to hold in the case of barium (Atomic No. =56) where very feeble rays were observed. This was also the case with iodine (Atomic No.=53) but no visible rays were obtained with samarium (Atomic No.=62).The vari- ation in intensity of the corpuscular rays excited by a radiation of frequency v is calculated; the total energy of the corpuscles composing the ray is AvCa”-Kc and if v is replaced by its value as a function of the atomic number the expression may be regarded as such a function. Brillouin’s theory (Compt. rend. 1920 170 274) introducing the idea of thermal agitation on electronic impact is briefly discussed and it is stated that its developments are not inconsistent with the experimental results obtained with respect to the photoelectric effect of X-rays. Dielectric Constants at the Critical Temperature. W. HERZ (2. physikal. Chem. 1922 103 269-272).-The author has calculated the dielectric constant of thirty-three liquids at the critical temperature.These substances a t ordinary temperature have dielectric constants varying between 81.1 and 1.491. The dielectric constant a t the critical temperature ought theoretically to be constant for all substances. The author finds that although the value is not constant it shows a decided tendency toward a constant value; the calculated values vary between 2-40 and 1-17. v4 H. J. E. J. F. S. The E and I Potential at the Interface Barium Sulphate- Water. A. GYEMANT (2 physikal. Chem. 1922 103 260- 268).-The dependence of the potential of barium sulphate onGENERAL AND PHYSICAL CHEMISTRY. ii. 53 the concentration of the barium-ions has been determined by means of E.M.P. measurements. The experiments lead to the conclusion that in order to explain the y potential the barium-ions and all other ions present must be taken into account.The observed dependence of the potential on the concentration of barium-ions cannot be deduced from the dependence of the E potential on the barium-ion concentration when the changes of E are evaluated by the method of Chapman (Phil. $lug. 1913 [vil 25 475) and of Herzfeld (Physikal. Z . 1920 21 28) in which the adsorption is neglected. J. F. S. Curves of Electrocapillarity in Non-aqueous Solutions. H. WILD (2. physikul. Chem. 1922,103 1-38).-Electrocapillarity curves have been constructed for saturated solutions of potassium chloride potassium nitrate mercurous nitrate and also for nitric acid in each of the two liquid layers of the liquid pairs water and amyl alcohol phenol furfuraldehyde ethyl acetate ethyl ether aniline chloroform propyl alcohol and isobutyl alcohol re- spectively the whole system in each case being in partition equilib- rium.The results show that two such solutions in equilibrium have the same Nernst potential rI=rr2 towards a given metal. This equality is true within the limits with which the absolute potential can be deduced from the electrocapillarity curves that is 5-10 m.-volts. With the same accuracy it is deduced that the tension at the interface between the two phases a t equilibrium r3 is zero or smaller than 5-10 m.-volts. This behaviour is true for all the cases examined and can be held to be a general rule for all solvents but the generalisation of the rule to all ions would be unsafe for it is quite possible that the capillary active ions may behave differently in different cases particularly in those cases where well-marked ion adsorption potentials are shown.I n these cases an additional potential difference must be noticed namely that due to the partition of the ions. J. F. S. Forces at the Boundary between Phases. EMIL BAUER (2. physikul. Chem. 1922 103 39-42).-A theoretical paper in which on the basis of the experiments of Wild (cf. preceding abstract) the author discusses the origin of the potential difference at the boundary between two liquid phases. In an earlier paper (A. 1916 ii 231) the author expressed the view that the potential difference was due to ion adsorption whilst Beutner (A. 1918 ii 214; 1919 ii 263) maintains that a partition of the ions between the phases is the cause.The author now shows that his view is suppbrted by the electrocapillarity measurements of Wild. J. F. S. Theory of Electrocapillarity. I. Electrocapillary Pheno- mena in Non-aqueous Solvents. A. FRUMKIN (2. physikul. Chem. 1922 103 43-54) .-Curves of electrocapillarity have been determined for 0.1N- and ,!-ammonium nitrate AT-sodium bromide and N-sodium iodide in methyl alcohol 0-2N-ammonium 3*ii. 54 ABSTRACTS OF CHEMICAL PAPERS. nitrate N-lithium chloride and N-sodium iodide in ethyl alcohol N-lithium chloride in mixtures of ethyl alcohol and water 0-9N- lithium nitrate 0-5N-lithium chloride and N-ammonium thio- cyanate in acetone N-ammonium thiocyanate and 1.6N-sodium iodide in pyridine. The results show that the activity of the anion in these solutions manifests itself in exactly the same way as in aqueous solutions. The maximum in these solutions when compared with aqueous solutions of corresponding composition is found to be displaced to the left that is it corresponds with a smaller cathodic polarisation.Theory of Electrocapillarity. 11. A. FRUMKIN (2. physikal. Chem. 1922 103 55-70; cf. preceding abstract).-It is shown that the whole of the electrocapillary phenomena are governed by the equation dy=cd+-Srldp where y is the thermodynamic potential in the solution containing pi ions (b is the potential difference of solution/metal and ri the number of ions which must be added to the solution so that pa remains constant when the surface of the metal is increased by a unit. The values of dy/dq and E the quantity of electricity combined with Fr ions have been determined experimentally for 2N-sulphuric acid saturated with mercurous sulphate E (calc.) 39 X coul./cm.2 E (obs.) 39 x 10-6 coul./cm2. ; N-sodium chloride saturated with mercurous chloride E (calc.) 50 x 10-6 coul./cm2. E (obs.) 47 x coul. /cm2. ; N-potassium hydroxide saturated with mercuric oxide E (calc.) 21 x 10-6 coul./cm2. E (obs.) 17 x coul./cm2. ; N-potass- ium nitrate and 0-OW-potassium iodide saturated with mercurous iodide E (calc.) 86 x 10-6 coul./cm.2 E (obs.) 90 x coul. /cm.2 with the results stated. It is shown that the Lippmann-Helm- holtz differential equation for solutions is true both for those with normal electrocapillarity curves and for those with abnormal curves. The potential of a dropping electrode always coiiicides with the value given by the corresponding electrocapillarity curve.A potential difference exists between mercury on the one hand and water methyl alcohol ethyl alcohol and acetone on the other when the surface layer contains either ions or adsorbed molecules of a dissolved substance. I n investigations of the present type it is necessary to differentiate between thermodynamic and electro- capillary solution tensions. J. F. S. Decomposition Tensions of Fused Mixtures of Sodium Hydroxide with Zinc Oxide or Cadmium Oxide. L. ROLLA and R. SALANI (Gaxxetta 1922 52 286-313).-Experiments similar to those of Sacher (A. 1902 ii 120) have been made on mixtures of molten sodium hydroxide with zinc oxide or cadmium oxide. Fused sodium hydroxide shows two points of cathodic decomposition the lower one 1-20 volts corresponding with the discharge potential of the hydrogen-ions ; the higher point corre- sponding with the discharge potential of the sodium-ions has the value 2-08 volts a t 460° 2.13 volts a t 412" and 2-24 volts at 364".With the mixtures containing zinc or cadmium oxide the decom- J. F. S.GENERAL AND PHYSICAL CHEMISTRY. ii. 55 position occurring is that of zinc or cadmium hydroxide the decom- position tensions for the zinc- and cadmium-ions being respectively 1.80 and 0-89 volts. T. H. P. The Electrolytic Dissociation of Dibasic Acids. Deter- mination of the Second Dissociation Constant of Acids from Electrometric Measurements. ERIK LARSSON (2. anorg. Chem. 1922 125 281-294).-The author applies the theory of Bjerrum (A.1919 ii 9) to the ionic equilibria in a solution which contains the neutral salt and the free dibasic acid. He shows how the second dissociation constant can be calculated from the hydrogen- ion activity measured electrometrically. The results obtained agree well with some earlier results obtained by the conductivity met hod. W. T. Formic Acid. 11. Electrolytic Dissociation of Formic Acid. FR. AUERBACH and H. ZEGLIN (2. physihZ. Chern. 1922 103 178-199).-The electrical conductivity of formic acid and of sodium formate has been determined over a wide range of con- centration at 18". It is shown in agreement with earlier investig- ators that small quantities of formic acid and sodium formate are decomposed a t the platinised electrodes.The decomposition has been investigated and in the case of the formate shown to consist in an oxidation to sodium hydrogen carbonate and to be due to the oxygen occluded in the platinised electrodes. This disturbing factor may be removed by suitable treatment of the electrodes with hydrogen before the measurements. The limiting value of the molecular conductivity of sodium formate at 18" is extrapolated to 91 and from this value the limiting value for formic acid is calculated to be 362.5. The dissociation constant for formic acid is shown to be inconstant and to vary with increasing dilution from 2-05 x lob4 to 1.91 x and this difference is shown not to be due to experimental error. Formic acid belongs there- fore to the acids of medium strength which do not follow the law of mass action closely.Values have been calculated for the electrical conductivity of formic acid which agree with the experi- mental values exactly over the whole range of concentration on the basis of Ghosh's hypothesis. The Anomaly of Strong Electrolytes. HENRY J. S. SAND (Phil. Mag. 1923 [vi] 44 129-144).-A critical examination of the theories of Ghosh (A. 1918 ii 215 348 392 790) and of Milner (A. 1918 ii 148). Employing the Boltzmann theorem and the Born theory of the potential energy of a pair of attracting ions estimations are made of the degree of association of a com- pletely ionised electrolyte immersed in a medium of uniform dielectric constant. It is shown that the probability of the two ions of a binary electrolyte (N/10-solutions) being in contact in a given small volume is only 8.1 times as great as the probability of their occurrence in any two volumes of the same size so situated that the attraction between the ions is negligible.The hypothesis of complete ionisation in the case of salts like sodium chloride is J. F. S. 3*-2ii. 56 ABSTRACTS OF CHEMICAL PAPERS. thus completely established. It is shown that a preponderating proportion of the ions will be subject to the inverse square law of electrical action. The relation P V = 2RT - +RThf( h) obtained by Milner from the Clausius' virial theorem can have only approximate validity. The Ghosh theory of the "crystalline" arrangement of the ions in aqueous solutions is criticised and i t is concluded that the space lattices in this theory are merely theoretical " distributions of reference." The deductions from Ghosh's and Milner's osmotic pressure formulae are compared with the experimental results.The agreement between the mean experimental values of (2-4) for univalent binary chlorides in aqueous solutions and the same value derived from Ghosh's formula is good ; the agreement a t the higher concentrations is better than that attained by Milner's theory. Relationship between the Specific Heat of Liquids. W. E. G. W. HERZ (2. anorg. Chem. 1922 125 295-300).-A theoretical paper. It is shown that the specific heats of liquids a t two-thirds their critical temperature divided by the values at one-half the critical temperature gives a fairly constant value. This constant for the twenty organic liquids quoted is about 0.8.Liquids in a homo- logous series show an increase of about 9 for the introduction of a CH group but with aniline-dimethylaniline there is an exception the difference being very small ; in this series the specific heat decreases with increasing molecular weight. In homologous series the molecular heat of evaporation increases about 10 for each CH group but there are many exceptions. The introduction of a chlonne atom in place of a hydrogen atom increases the molecular heat of evaporation by about 9.5 units the introduction of a second chlorine atom has a less effect. A double bond lowers it by about two units. W. T. [Determination of Boiling Points.] Boiling Points of Ammonia Sulphur Dioxide and Nitrous Oxide.F. W. BERGSTROM ( J . Physical Chem. 192Z726,876-894).-A comparison has been made of the efficiency of single-walled and vacuum- jacketed boiling vessels. It is found that the boiling point of ammonia determined in a single-walled vessel is about 0-2" high even with internal electrical heating. Determination in a Cottrell tube reduces this error. Boiling points may be determined accur- ately in a vacuum-jacketed vessel a t temperatures as low as - 90° but in such cases a correction should be applied for the depth of immersion of the bulb and there should be no uncooled stem. The Cottrell apparatus can be used successfully a t temperatures at least as low as - 33.4". Better equilibrium between vapour and liquid is obtained and uncertain temperature corrections are eliminated by using this apparatus and a smaller quantity of liquid is generally required.The following boiling points have been determined a t 760 mm. ammonia - 33-41"&0~1" sulphur dioxide - 10~02"&-0~1" and nitrous oxide - 89.5"&0*2". J. F. S.GENERAL AND PHYSICAL CHEMISTRY. ii. 57 Application of the Method of Continuous Variations to Ebullioscopic Phenomena for the Determination of Double Salts in Solution. F. BOURION and E. ROUYER (Compt. rend. 1922 175 1406-1408).-The method of continuous variations as applied to ebullioscopic measurements was tested in the case of solutions of cadmium chloride with potassium chloride and ammonium chloride respectively and solutions of cadmium iodide with potassium iodide and found to give quite as satisfactory results as when applied to cryoscopic measurements.W. G. A Micro-method for the Determination of Molecular Weight in a Melting-point Apparatus. 11. Determinations with Extremely Minute Quantities. KARL RAST (Ber. 1922 55 [B] 3727-372S).-The author has succeeded in further refining his micro-method for the determination of molecular weight in freezing camphor to such an extent that it is possible t o obtain accurate results with scarcely visible amounts of substance. The capillary tube is slightly conical in shape and rather wider (2-3 mm.) than those recommended previously; it is essential that i t should be very thin in the wall and that the bottom should be hemispherical. The solutions are prepared in the capillary itself the substance under investigation being first introduced and sub- sequently the camphor.The materials are pressed together by a small glass rod. The capillary is sealed and subsequently drawn out to a long thread by means of which i t is attached to the thermometer. Mixing of the components is effected by melting and re-solidification. The column in the capillary must not be more than 2 mm. in height so that in general 0.2-0.3 mg. of substance and 2-3 mg. of camphor are required. It is essential to guard against undue concentration of the solutions which however may sometimes be greater than normal. H. W. The Influence of the Velocity of Cooling on the Hardness and Microstructure of Eutectic Mixtures. N. S. KURNAKOV and A. N. ACHNASAROV (Z. unorg. Chem. 1922 125 185-206).- The hardness of eutectic mixtures increases with the rate of cooling and with the degree of fineness of the grains. This tendency to increase in hardness decreases with increasing brittleness e.g.zinc-antimony. Increasing fineness of grains also results in increasing passivity of the alloy. Homogeneous phases of pure metal and solid solutions show no change in hardness on rapid cooling. The above conclusions were obtained from a study of the systems cadmium-silver silver-copper gold-nickel zinc- antimony. W. T. General Theory of the Adsorption of Solutions. BROR. GUSTAVER (Kolloid Z. 1922 31 358-362).-A theoretical paper in which the author criticises the views put forward on the adsorp- tion of solutions by Ostwald and Izaguirre (A. 1922 ii 480). It is shown that the theory is not in keeping with the author's results on the sorption of vapours by charcoal (A.1922 ii 479) neither isii. 58 ABSTRACTS OF CHEMICAL PAPERS. it in keeping with respect to the thickness of the adsorbed layer as determined by the author and others. J. F. S. Adsorption of Ions by Freshly Precipitated Manganese Dioxide. P. B. GANGULI and N. R. DHAR ( J . Physical Chem. 1922 26 836-844) .-The authors have investigated the adsorp- tion of kations by manganese dioxide and also the effect of various anions on the adsorption. Manganese dioxide was prepared in the solution of ions under investigation by the addition of equiva- lent quantities of potassium permanganate and manganous sulphate and the amount of adsorption determined by analysing the filtered solution after equilibrium had been reached. Some thirty-five salts have been used in the investigation and the results show that the coagulating powers of the different electrolytes as calculated from the percentage of kation adsorbed from approximately normal solutions of the electrolytes follow the Schulze-Hardy law very imperfectly.The effect of the anions on the adsorption of kations by manganese dioxide is found to be very marked. There is however no regularity in the variations shown by the adsorption values of the kations with variation of the anion. In the case of ferric salts the adsorption of ferric-ions by manganese dioxide is abnormally large; a result which is probably to be attributed to a partial hydrolysis of the ferric salt with the separation of ferric hydroxide.Among the electrolytes of metals occurring in the same group of the periodic system the values of the percentage adsorption are generally found to be in the order of the atomic weights of the kations. Physical Chemistry of Dyeing. Acid and Basic Dyes. T. R . BRIGGS and ARTHUR W. BULL ( J . Physical Chem. 1922 26 845-875).-The process of dyeing wool with acidic and basic dyes has been investigated from the point of view of the adsorp- tion hypothesis as formulated by Pelet- Jolivet and Bancroft (Applied Colloid Chemistry 1921 115). The effect of dyes on the adsorption of acids by wool and of acids on the adsorption of dyes has been determined quantitatively for typical acid dyes. It is shown that the taking up of dyes is a case of adsorption and that the amount of dye adsorbed varies continuously with a change in the hydrogen-ion concentration of the dye-bath.No evidence of chemical action between dyes and wool has been obtained. J. F. S. J. F. S. The Determination of the Dissociation Pressures of Hydr- ated Salts by a Dynamical Method. 11. JAMES RIDDICK PARTINGTON and DONALD BENNETT HUNTINGFORD (T. 1923 123 160-170). A New Explanation of Diffusion. ALEXANDR BAT~K (Chem. Listy 16 [9] 295-299).-Pick’s theory for the diffusion of liquids (Ann. Phys. Chem. 1855 [ii] 94 59) is discussed and its experimental basis questioned. Certain discrepancies betweenGENERAL AND PHYSICAL CHEMISTRY. ii. 59 the results calculated on this theory and those obtained experi- mentally by Voit (Ann. Phys. Chem. 1867 [ii] 130) are pointed out.The technique of optical methods for the determination of rates of diffusion is examined and possible sources of error are suggested. Weber's confirmation (Awn. Phys. Chem. 1870 [iii] 7 469 536) of the applicability of Fourier's law to the diffusion of liquids is shown to be doubtful and the necessity for a recon- sideration of the whole question emphasised. The mechanism of the diffusion of liquids is then discussed from first principles and the use of the conception of limiting states of solutions in helping to elucidate the problems of diffusion of liquids is substantiated. R. T. Diffusion in Solid Solutions. H. Wmss and P. WERRY (Compt. rend. 1922 175 1402-1405).-A study of the inter- penetration by diffusion of gold and silver at the temperatures 935" 885" and 835" shows that the form of the law of diffusion in fluids is valid.The values of the constant K a t the temperatures used when interpolated to 870° give a value 0.0000375 which is in very close agreement with the value obtained by Fraenkel and Houben (A. 1921 ii 491). Process of Diffusion in Gelatin. Liesegang's Ph&omenon. CARL ADOLF SCHLEUSSNER (Kolloid Z. 1922 31 347-352).- With the object of employing the diffusion into gelatin as a means of testing the suitability of various specimens of this substance for use in the manufacture of photographic plates the author has examined the conditions under which the Liesegang rings are produced. It is shown that for a regular and uniform diffusion the gelatin must be in a uniform condition. The necessary condition can only be obtained after the gelatin has been held in solution for at least twenty-four hours.It is shown that from the character of the ring formation conclusions may be drawn as to the purity of the gelatin. The intermediate rings which may be observed with a lens between the main rings are shown to be silver salts of phosphoric and halogen acids and are due to impurities in the gelatin. A bibliography of the work on the formation of Liesegang rings and allied phenomena is included in the paper. The Relation between the Crystal Structure and the Con- stitution of Carbon Compounds. I. Compounds of the Type CX,. Racemates and Pseudo-racemates. PHILIPPE LANDRIEU (Bull. Xoc. chim. W. G. J. F. S. ISABEL ELLIE KNAGGS (T. 1923 123 71-79). Double Compounds and Mixed Crystals.1922 [iv] 31 1217-1241).-Lectures delivered a t the College de France. G. F. M. Hydrogen-ion Concentration and the Properties of Emulsoid Colloids. ROBERT HERMAN BOGUE ( J . Physical Chem. 1922 26 801-811).-A theoretical paper in which it is shown that the various physical properties of the emulsoid colloids includingii. 60 ABSTRACTS OF CHEMICAL PAPERS. the viscosity jelly strength melting point and joining strength are a t a minimum a t a hydrogen-ion concentration corresponding with the isoelectric point. As the acidity or alkalinity of the solution is increased from this point the whole of these properties increase in value. It is shown that salt precipitations for gelatin contents should be made a t the isoelectric point if the maximum precipitation is to be obtained.The necessity for a careful control! of the hydrogen-ion concentration in investigations on the proteins is emphasised and the desirability of a similar control in the gelatin and glue plant during manufacture is also pointed out. The limitation of the benefit obtainable from such conDro1 however makes the practicability of such methods when applied to the improvement of the quality of the material very questionable. The estimation of the hydrogen-ion concentration is urged as a test in the evaluation of gelatin and glue but it is not recommended that all tests of viscosity jelly strength and joining strength be made a t a specified hydrogen-ion concentration. J. F S. Present Position of the Theory of Peptisation. W. MOELLER (2. Leder Gerb.Chem. 1922 1 360-376; cf. A. 1915 i 439).- A theoretical paper in which the author discusses the theory of peptisatiod particularly in connexion with its application to tanning and the leather industry. It is shown that the mechanism of tanning consists in the destruction of the sol-condition of the peptised solution by the substance of the hide. The hide substance removes the sol by adsorbing the peptising agent and the separ- ated coagulated oil surrounds the micells of hide to form micro- crystals. The author defines tan as a system which of itself is quite insoluble but by peptisation may be converted into a colloidal solution. Leather is defined as an animal hide the elementary particles of which are crystalline micells protected by a sheath of tan particles from hydrolytic influences.J. F. S. Protective Colloids. XII. Gelatin as a Protective Colloid. 111. Colloidal Platinum. A. GUTBIER and A. ZWEIGLE (Kolloid Z. 1922 31 346-347; cf. A. 1922 ii 485).-Solid platinum colloids may be prepared by reducing chloroplatinic acid in aqueous chloroform solution by means of hydrazine hydrate in the presence of solutions of gelatin. When five parts of a 1 yo solution of chloro- platinic acid in water saturated with chloroform is mixed with five parts of o*14y0 gelatin solution and slowly reduced by the addition of 0.1 yo solution of hydrazine hydrate a colloidal solution which is black by transmitted light and dark brown by reflected light is produced. This sol on keeping deposits a small quantity of a black powder and on dialysis yields a very stable sol.The black precipitate is reversible t o the extent of about 70%. The sol is stable towards heat and quite insensitive to electrolytes which are capable of producing an hydroxyl-ion. Solutions of neutral salts coagulate the sol only after keeping for several days whilst acids even in dilute solution produce a rapid coagulation. On evaporation a t 27" in a vacuum over sulphuric acid a solidGENERAL -4ND PHYSICAL CHEMISTRY. ii. 61 colloid is obtained which is completely reversible in cold water and contains 18*76y0 of platinum. When the quantities of the reagents employed are one part of 0.14% gelatin five parts of 0.1 yo solution of chloroplatinic acid and four parts of chloroform water the solid sol obtained by the same process is also com- pletely reversible to form stable secondary sols in dilute solution and the solid itself contains 55-67y0 of platinum.J. E. S. Validity of the Law of Mass Action for Ionic Equilibria. J. N. BRONSTED and KAI PEDERSEN (2. physikal. Chem. 1922 103,307-31 5) .-The reaction between ferric chloride and potassium iodide has been investigated in aqueous solution a t 25" and from the results it is shown that the law of mass action is valid for ionic equilibria when these are set up in concentrated salt solutions as solvents. A convenient method for determining solubility in the absence of air is described and used to determine the dissociation constant of the tri-iodide-ion. The value for the equilibrium constant of the dissociation IiZZI,+I' a t 25" is 0*00611 a t 15*0" 0.00502 and a t 18.5" OaO0540.J. P. S. Formic Acid. 111. Partition of Formic Acid between Ether and Water and its Application to Analytical Purposes. FR. AuERBacH and H. ZEGLIN (2. physikaE. Chem. 1922 103 200-237).-The partition coefficient of formic acid has been determined a t 18" for ether and water O*SN-sulphuric acid and 0*5N-sulphuric acid containing 100 g. of sodium chloride per litre respectively. In the case of ether and water the coefficient was determined for concentrations up to 1 mol. per litre and in spite of corrections for the electrolytic dissociation of the formic acid in water the partition coefficient of the undissociated acid was found to be inconstant but to vary in a linear manner with the con- centration in the aqueous phase ; the value f =C,/Cw varies from 0.395 for Cw=0*0446 to 0.454 for C,=1*343.In the case of ether and 0-SN-sulphuric acid in which the electrolytic dissociation in the aqueous phase is practically zero the partition is displaced in favour of the ether layer but here also the coefficient is dependent on the concentration of the aqueous layer. Sulphuric acid con- taining sodium chloride gave similar results to the acid and ether alone. The progression of the partition coefficient is explained by the polymerisation of the formic acid in ethereal solution From the law of mass action the extent of the polymerisation is calculated on the assumption of double molecules by means of the formula x=[(HCO,H),]/[HCO,H]~ ; the following values of x are obtained for ether saturated with water x=0.139; for ether saturated with 0.5N-sulphuric acid x=0-147 and for ether satur- ated with 0-SN-sulphuric acid containing 100 g.of sodium chloride per litre x=0*158. The small differences in the polymerisation constant are due to the varying water content of the aqueous phase. The true partition coefficients have been calculated from the above-mentioned quantities and the experimental results. The partition coefficient of undissociated unimolecular formic acidii. 62 ABSTRACTS OF CHEMICAL PAPERS. a t 18" has the following values between ether and water,fo=0*393 ; between ether and 0.5N-sulphuric acid f0=0*398 and between ether and O.5N-sulphuric acid containing 100 g . of sodium chloride per litre f0=0*400. For the purpose of applying the partition coefficient to analytical purposes the volume changes of the phases when ether is shaken with water or N/S-sulphuric acid solution were either determined or taken from the unpublished work of Mylius.By means of the volume correction factor and the par- tition coefficient the concentration of formic acid in a 0.5N-aqueous sulphuric acid solution may be calculated from the concentration in the ethereal layer after shaking the sulphuric acid solution with ether. Equations for this calculation have been derived and tables of numerical factors are given in the paper. Formic acid in 0*5N-sulphuric acid solutions may be estimated by this method for amounts of formic acid between 2.4 g. and 0.011 g. with an average error of 0*38%. If the sulphuric acid also contains 100 g.of sodium chloride per litre the average error of the estimations is 0.26%. J. P. S. J. A. CHRISTIAN- SEN (8. physikal. Chem. 1922 103 91-98).-A theoretical paper in which on the assumption that only molecules in a definite quantic condition react the author has developed a velocity formula for reactions of the type AB + A +B. The equation deduced has the form E= (pm/pn)e-(em-+JIRF . v'/( 1 -e-hv'/R*) where E is the velocity constant pm and pn are the a prior! probabilities respectively that the molecules are in the m-quantlc or normal condition when the m-quantic condition is that necessary for reaction ; the other symbols have their usual significance; those with the dash such as v' refer to the products of the reaction. The author shows that when hv'/RT is small the equation reduces to the form k=(pm/pn) x e-(em-cn)lRF - RT/h which is practically the same as the expression deduced by Herzfeld (A.1922 ii 136) from Stern's expression for unimolecular reactions. When hv' /RT has a comparatively large value the expression reduces to k= (pm/pn) . e-('m-en)/RT x v' which is the same as the expression obtained by Dushman (A. 1921 ii 315). J. F. S. Thermal Decomposition of Carbonyl Chloride. J. A. CHRISTIANSEN (8. physikal. Chem. 1922 103 99-138).-The thermal decomposition of carbonyl chloride both alone and mixed with chlorine has been investigated' at a number of temperatures between 655" and 782" absolute. The results indicate that over the range of temperature 705-745" Abs. the reaction takes place according to the equation h=k~Cc,,(Ccoaz-f) where his the velocity k a constant and f=Cco .C,,/K K is the equilibrium constant of the reaction COC1,t CO+Cl,. The equation only holds when the concentration of the carbonyl chloride carbon monoxide and chlorine are of the same order (10-3-10-4 mol./litre). At the commencement of the reaction that is until the above condition obtains the divergences from the equation are always positive Velocity Law of Unimolecular Reactions.GENERAL AND PHYSICAL CHEMISTRY. ii. 63 in the sense that the observed velocity is always greater than the value calculated according to the above equation. The velocity is not markedly affected by a change in the relative glass surface in the ratio 50138. The velocity is unaffected by the presence of air when the concentrations of carbon monoxide and chlorine are considerable from the commencement of the reaction.When it is assumed that the expression for the velocity does not change with temperature the value of the constant k can be found by means of the expression log k= -11420/T+15.154 for the temperature range 685-782". An experiment a t 655" was found to be less exact but it agreed with the above expression within the limita of the experimental error. The Velocity of Reaction in Mixed Solvents. IV. The Influence of the Base on the Velocity of Saponification of Esters. ALBERT ERIC CASHMORE HAMILTON MCCOMBIE and HAROLD ARCHIBALD SCARBOROUGH (T. 1923 123 197-207). Chemical Kinetics of Heterogeneous Systems. IV. The Mechanism of Chemical Reaction when Noble Metals Dis- solve in Potassium Cyanide Solution.EIICHI YAMAZAKI ( J . Chem. SOC. Japan 1922 43 686-690).-According to Bodlander (2. anorg. Chern. 1896 19 583) gold and other noble metals dissolve in potassium cyanide solution owing to the catalytic action of hydrogen peroxide produced as an intermediate; but this explanation and Bodlander's formulz cannot explain the author's experiences. The reaction velocity is approximately pro- portional to the square root of the concentration of oxygen but not to the concentration itself. The production of hydrogen peroxide is a secondary reaction which may of course accelerate the dissolution of the metals and produce the peroxides when Ba" or Ca" are present in the solution. When the dissolution of the metal is normal the reaction proceeds as follows Ag=Ag'+O ; 0+2@-+0" ; H20+0"+20H' ; or 2Ag+H,O+30,=2Ago+ 20H' etc.K. K. $ome Properties of the Active Nickel used as Catalyst in Organic Chemistry. ANDRB BROCHET (Compt. rend. 1922 175 1073-1075 ; cf. this vol. ii 181.-The pyrophoric property of nickel prepared for use as a catalyst is entirely due to occluded hydrogen and if the metal is freed from that gas under conditions which maintain in the nickel the property of being readily trans- formed into oxide its catalytic activity remains unimpaired. There is no relation between the pyrophoric condition and the catalytic efficiency although the extent of the former is a convenient indi- cation of the progress of the preparation of the catalyst. Catalytic nickel if washed and dried loses its pyrophoric property; in cases in which the catalytic activity is affected by this operation it may be restored by heating for a short time in hydrogen. The author appends a brief discussion from the point of view of employment as a catalyst of the extent and condition of the metallic surface and the property of occluding hydrogen.J. F. S. H. J. E.ii. 64 ABSTRAOTS OF CHEMICAL PAPERS. Catalytic Hydrogenation with Nickel. Factors Determin- ing Catalytic Activity. R. THOMAS ( J . SOC. Chem. In&. 1923 42 21-26~).-A description is given of some of the methods which may be employed for the preparation of catalytically active nickel and the relative activities of variously prepared nickel catalysts in the hydrogenation of oils have been determined.The most active catalyst was that obtained by the reduction by means of hydrogen a t 350-500" of nickel hydroxide precipitated on kieselguhr. A nickel catalyst prepared by reduction with charcoal a t 600° or by electrical disintegration or by the decomposition of nickel carbonyl was less than one-fourth as active whilst nickel precipitated from its salts by means of aluminium showed only about one-sixth of the activity. The author concludes with a theoretical discussion as to what are ultimate factors which influence catalytic activity and what form of energy transference occurs between the catalyst and reacting substances. It is suggested that the catalyst activates a certain number of molecules of the reactants by thermionic emission and that the bulk of the mole- cules are then successively activated by electron emission from molecules already in an activated condition through the instru- mentality of the catalyst.G. F. M. Heterogeneous Catalysis and the Orientation of Adsorbed Molecules. H. R. KRUYT and C. F. VAN DUIN (Proc. K . Alcad. Wetensch. Amsterdam 1923 25 324-326) .-In continuation of previous work (A. 1921 ii 392) the authors find that the reaction between aqueous solutions of dibromosuccinic acid (meso- or racemic form) and potassium iodide is considerably accelerated in the presence of charcoal. The greater acceleration occurred in the case of the meso-form a result in accord with stereochemical considerations. The result confirms the conclusion that positive contact catalysis is to be anticipated only when the reacting group of molecules is directed away from the adsorbent and towards the surrounding medium.J. S. G. T. A Type of Ideal Electric Atoms. J. L. (Nature 1922 110 873).-The mathematical solutions arrived at by Hargreaves (this vol. ii 19) give a possible structure for an ideal atomic nucleus of the Rutherford type. Speculation Concerning the Positive Electron. HORACE H. POOLE (Nature 1923 111 15-16).-Lodgeys speculation (A. 1922 ii 836) is examined from the point of view of the relative abundance of the lighter and heavier elements the case of hydrogen being considered in particular. An immense discrepancy is appar- ent. Doubtless some negative protons would combine with previously formed positive complexes but since about half the complex nuclei first formed would be negative so that some of the positive protons would be lost by combination with them these effects would be expected to balance approximately.If however it is assumed that when two unequal nuclei combine the A. A. E.INORGANIC CHEMISTRY. ii. 65 sign of the combination is determined by that of the larger cornti- tuent it is conceivable that if the first set of nuclei formed happened to be positive they might so direct the course of subsequent events as to lead to the existing distribution of the elements. A Laboratory Apparatus for Rapidly Drying Sensitive Organic Substances. J. BOUILLOT ( J . Phclrm. Chim. 1923 [vii] 27 23-28).-A tube of glass 15 cm. long and 30 mm. wide sealed at one end and fitted with a heat-resistant cork at the other carries the siibstance to be dried in a boat The tube has two slightly narrower tubes attached vertically t,o its upper surface by narrow necks contlaining plugs of cotton wool and in the openings of these tubes are fitted corks carrying narrow glass tubing termin- ated inside by being drawn out in narrow hooks so as to avoid violent currents of air impinging on the boat below. The whole apparatus can be immersed in an air-bath at the required tem- perature and currents of dried air can be aspirated over the contents of the boat a t any required pressure. J. F. BREWSTER ( J . Ind. Eng. Chem. 1923 15 32).-A solid rubber stopper is cut about three-quarters through a t the small end so that a thin flap of rubber is left to serve as the valve leaf. The stopper is then bored leaving the flap intact to receive a short length of glass tubing which is pushed through to within a few mm. of the valve seating. A piece of glass tubing of such diameter as to allow free play of the valve flap but at the same time to fit the stoppers tightly is closed a t one end by the valve stopper and a t the other by a one-hole stopper. The contrivance will allow the passage of air or liquid from the direction of the valve stopper to the one-hole stopper but not in the reverse direction and is useful for preventing amongst other things a " suck back " of wafer from a water vacuum pump into the apparatus during a vacuum distillation etc. A. A. E. H. K. A Simple Check Valve. G. F. M.