ii 81 General and Physical Chemistry. Refractive Index of Gas Mixtures. S. VALENTINER and 0. ZIMMER (Bar. Deut. phys~kal. Ges. 1913. 15 13()1-131ti).-The authors have measured. the refractive indices of mixtures contain- ing hydrogen and carbon dioxide and also hydrogen helium and neon. I n the latter mixtures the proportion of helium to neon was constant a mixture of these two gases containing 27.8% of helium being used in the preparation of the gases submitted to investiga- tion. The refractivity data were obtained with the aid of a Zeiss interferometer. The results indicate that the refractivity of mixtures of hydrogen and carbon dioxide is slightly less than that calculated according t o the mixture rule. I n the case of the mixtures of hydrogen helium and neon the deviations from the mixture rule values are within the limits of possible experimental error.H. M. D. Mixed Liquid Crystals. PAUL GAUBERT (Compt. rend. 191 3 157 1446-1448. Compare A. 1911 ii 949; 1912 i i 109 510).-'l'he author has extended his study of the variation in birefraction of an optically negative crystalline liquid on addition of a substance giving optically positive liquid crystals to mixtures of substances capable of mixing in all proportions. The two substances studied were cholesteryl propionate and ethyl anisylideneaminocinnamate the results are shown in a table and graphs the measurements being made'at looo f o r the rays B C D 3 P. The rotatory power is found tot increase considerably with the birefraction of the mixture. W. G. Spectroheliographic Investigation of the Arc Light.OTTO OLDENBERG (Zeitsch. wias. Photochem. 191.3 13 133-172).-8n apparatus is described by means of wllich the author has obtained monochromatic reproductions of the arc light on photographic plates. By impregnating the carbons with metallic salts it was possible to examine the behaviour of well-known series lineB and bands. In the case of sodium the photographic records show that the principal and second subordinate series of lines are mainly emitted by the periphera.1 layers of the arc whilst the first subordinate series is due to emission from layers nearer the axis. Lithium behaves similarly to sodium but the observations with aluminium magnesium and calcium indicate that these behave quite differently from the alkali metals.The monochromatic reproductions show further that there are two distinct types of bands. I n the case of the carbon and cyanogen bands the intensity diminishes from the centre to t h s edge of the arc whilst all the other bands examined are characterised by greater intensity at the edge of the arc. I n view of the very VOL. CVI. ii. 6ii. 82 ABSTRACTS OF CHEMICAL PAPERS. considerable difference between the monochromatic pliotograplis of bands and lines it is suggested that this method may be made use of in doubtful cases to determine the character of a particular emission. H. M. D. Influence of the Constituents of the Crystal on the Form of the Spectrum in the X-Ray Spectrometer. W. H. BRAGG (Proc. Roy. Soc. 1914 [A] 89 430-438).-1t has been observed that the relative intensities of the different parts of an X-ray spectrum may be altered very considerably by changing the crystal which is employed in the spectrometer and an account is given of experi- ments which have been made to determine the origin of this effect.According to the observations of Barkla every metal emits characteristic rays which can only be excited by rays character- istic of metals which are of higher atomic weight. These observa- tions have also shown that homogeneous X-rays are strongly absorbed by any substance in which they can excite the rays characteristic of that suhstance. As a consequence of these rela- tions the connexion between the atomic weight of an absorbing screen and its power of absorbing X-rays of given quality exhibits marked discontinuities and it is shown that the operation of this factor through the absorbing action of the atoms of which the crystal is composed gives rise to the variations in the relative intensities of the different parts of the spectrum.H. M. D. A Seoond Spectrum of Neon. THOMAS RALPH MERTON (Proc. Boy. Xoc. 1914 [A] 89 447-451).-A comparison has been made of the spectra obtained with a weak discharge from an induction coil and with the same discharge when a small Leyden jar and a spark-gap of about 2 mm. were introduced into the circuit. I n the mom refrangible region the jar-discharge spectrum was found to consist of a large number of lines which are not present in the ordinary neon spectrum whilst the strongest of the ordinary neon lines were only just visible.The wavelengths of most of these lines were measured from photographs taken with the grating spectrograph but some of the lines in the green are so feeble that it was found necessary to obtain photographic records with very long exposure on the prism spectrograph. The wave-length data agree with those recorded by Liveing and Dewar (A. 1901 ii 213) although about 50% of the Lines given by these authors have not been observed. The experiments indicate that neon like argon krypton and xenon is capable of giving two different spectra according to the manner in which it is electricdly excited. H. M. D. The Band Spectrum of Aluminium and its Presence in the Flame Spectra of Certain Minerals. ARNAUD DE GRAMONT (Compt. rend. 1913 157 1364-1368).-The author has discovered a sixth absorption band in the spectrum of aluminium much more refrangible and much weaker than the other five. It lies between h 4330.5 and h 4462.6 becoming degraded towards the red.Work-GENERAL AND PHYSICAL CHEMISTRY. ii. 83 ing with an oxy-acetylene flame the author finds that in the case of pure substances the true aluminium spectrum is only obtained when using the metal itself or its haloids the oxygen compounds not giving the spectrum as described. He has also studied a large number of minerals and examined them carefully for halogens and finds that for them the above rule does not hold good but that the spectrum is more probably controlled by the constitution of the mineral. W. G. Absorption of Inorganic Salts in tbe Ultra-violet. T. RETSCHINSKY (Ann.Yhysik 1913 [iv]. 42 1580-1588).-The ultra- violet absorption of lithium chloride lithium bromide sodium chloride sodium bromide sodium iodide and potassium iodide has been examined in aqueous solutions of varying concentration. All these salts show a strong absorption maximum in the extreme ultr% violet beyond h=220 pp. I n the case of lithium bromide there is also a weaker absorption band between 255 and 280 pp and some evidence of a further band a t about 300pp.. The former band is much less clearly marked in the cam of sodium bromide but both salts behave the same in regard t o the limiting wave-length of ultra-violet transmission. The chlorides and bromides conform to Beer's law but slight deviations have been found in the ca50 of the iodides in that the absorption increases with the concentration of the solution when the product of this concentration and of the thickness of the absorbing layer remains constant.The absorption curve is shifted towards greater wave-lengths when the temperature is raised to looo in the case of sodium chloride lithium bromide sodium bromide and potassium iodide but no shift is observed with lithium chloride. When the limiting absorption of the dissolved salts is compared with that found for the molten salts (A. 1908 ii 910) it is found that the straight lines which represent the influence of temperature in the case of the molten salts pass through the points which corre- spond with the limiting absorption in aqueous solution at 20° and looo. I n making this comparison the concentrations of the solutions were chosen so as to give the same number of absorbing molecules in the two cases.H. M. D. Absorption Spectra of Nickel Derivatives of Primary Aliphatic Nitroamines. A. P. N. FRANCHIMONT and H. J. BACKER (Rec. trav. chim. 1913 32 321-324. Compare T. 1912 101 2256 ; A. 1913 i 833).-'l'he nickel derivatives of methyl- and ethyl- nitroarnines have been prepared and their absorption spectra compared with that of nickel nitrate. The amine derivatives are prepared by treating freshly precipitated nickel hydroxide with aqueous solutions of methyl- and ethyl-nitroamines respectively. It is shown that the colour of the nickel nitroamines resembles that of nickel nitrate much more than the colour of the corresponding derivatives of copper and cobalt resembles that of their nitrates.Nickel methylnitroamine crystallises with 6H,O and possesses a bluish-green colour ; i t loses its water on remaining over sulphuric 6-2ii. 84 ABSTRACTS OF CHEMICAL PAPERS. acid and becomes intensely green. The anhydrous salt is more easily soluble in water than the crystallised salt. All three salta in N/Z-solution have maxima of absorption at l/h=2850 and ca. 1/h=2000 whilst t h e nitrate has a third band at 3810. The absorption curves for the methyl- and ethyl-nitroamine derivatives coincide over practically their whole length and the absorption is compleb in the ultra-violet. Spectrographic Examination of the Colour of Picrylalkyl- nitroarnines with Alkalis. A. P. N. FHANCHIMONT and H. J. BACKER (Rev. trwv. chim. 191 3 32 325-333).-The ahwrption spectra of alcoholic solutions of picrylmethylamine and picryl- methylnitroamine have been examined both alone and in the presence of potassium hydroxide.A solution N/5000 of picryl- methylamine exhibits two absorption bands with their heads a t 1/h=2390 and 2875; there is a maximum of transmission a t 2600 and a t 3490. On adding a slight excess of potassium hydroxide both bands persist but the head of the first is shifted to l/h=2420. Picrylmethylnitroamine exhibits general absorption but on the addition of alkali two bands appear a t 1/h=1975 and 2350. On allowing the alkaline solution to remain for twenty-four hours the character of the absorption spectrum is entirely changed and on measurement is found to be that of potassium picrate thus show- ing that picrylmethylnitroamine is hydrolysed by a slight excess of potassium hydroxide in cold solution according t o the equation C,H,(NO,),*NMe*NO + H,O = C,H,(NO,),-OH + NHMeONO,.N / 200-Solutions of picrylmethylamiiie and picrylmethylnitroamine have about the same colour but on dilution the colour of the picrylmethylamine is only slightly diminished whilst that of picryl- methylnitroamine e'ntirely disappears. Selective Absorption of Ketones. GEORGE GERALD HENDERSON and ISIDOR MORRIS HEILBRON (Proc. Roy. SOC. 1914 [AJ 89,414-418). -It is suggested that selective absorption of ultra-violet rays by ketones is caused by intramolecular vibrations due to the alternate forination and breaking down of unstable ring systems and that the momentary formation of these is effected through the agency of free partial valencies which make their appearance on the atoms of the compound under certain conditions.I n the case of acetone the two alternating phases may be represented by the formulz J. F. S. J. F. S. 1 ; CH~*C-CH -+ 1 f- O.-..-.H O H i . . and the selective absorption is attributed t o the' ele&ronic disturb- ances which accompany this intramolecular oscillation. The semicarbazones of acetone and its homologues exhibit only general absorption and this is also in agreement with the above hypothesis if it is assumed that the attraction between hydrogen and nitrogen is not sufficiently strong t o cause the appearance of free partial valencies.GENERAL AND PHYSICAL CHEMISTRY. ii. 85 The theory can be extended to other groups of ketones of which diacetyl [dimethyl diketone] and acetylacetone may be taken as examples.H. M. D. Absorption of the Ultra-violet Rays by Alkaloids of the Morphine Group and by Phenanthrene. M. GOMPEL and VICTOR BENRI (Compt. rend. 1913 157 1422-1425).-A study of the ultra-violet absorption spectra of phenanthrene morphine codeine and apomorphine with a view to throwing some light on their relative constitutions. The absorption curve of phenanthrene can be divided into three regions (1) between A 3800 and 3050 in which region E. varies between 200 and 600 and there are five absorption bands; (2) between A 2990 and 2700 where E. varies between 2000 and 10,000 and there are two absorption bands one strong (A 2925) and the other weak (A 2810); and (3) the region between ~ 2 7 0 0 and 2144 where E varies between 10,000 and 135,000 and there is a very intense band at ~ 2 5 0 3 .Morphine and codeine present almost identical absorption curves having one single broad absorption band a t h 2850 where ~.=1600. The curve then passes to a minimum a t h2620 and finally rises steadily to h 2195 where E. = 20,000. The curve for apomorphine resembles that of phenanthrene much more closely except that the intense band a t A2503 is missing. The authors therefore suggest that this band is produced by the double linking between the carbon atoms 9 and 10 this being missing in apomorphine. apollorphine can be detected spectrographically a t a dilution of 0.01 mg. per c.c. whilst for morphine and codeine the extract must contain a t least 0.1 mg.per C.C. W. G. Photographic Determination of the Fluorescence Spectra of the Chlorophyll Pigments. CH. DHBR~ (Compt. rend. 1914 158 64-67).-An examination of the fluorescence spectra of a- and P-chlorophyll carotin and the xanthophylls. Both a- and &chlorophyll show only one band of fluorescence the axis of which is very slightly displaced towards the red as compared with that of the corresponding absorption band this displacement being greater in the case of the f?- than the a-variety. I n the case of carotin a slight green fluorescence is obtained the band of which is distinct from ~ = 5 0 0 p p to 560 pp. The xanthophylls show no visible fluorescence. The examination was made of solutions of the pigments in anhydrous ethyl ether and photographs are given for a- and &chlorophyll.W. G. Infiuepce of Intramolecular Relative Vibration on the Intensity of the Absorption and Fluorescence of Valency Electrons. J. STARK and P. LIPP (Zeitsch. physikal. Chem. 1913 86 36-50).-A theoretical paper in which the general statement is deduced If of two substances A and B each containing similar atoms with similarly bound valency electrons the number of colli sions between the valency electrons with atom groups or otherii. 86 ABSTRACTS OF CHEMICAL PAPERS. molecules is greater in the case of B than that of A then the absorption bands of B due to the valency electrons will be stronger than those in the case of A and the fluorescence of A will be stronger than that of B. I n other words an increase in the number of collisions between valency electrons and atom groups or molecules increases the light absorption and weakens the fluores- cence.The fluorescence and absorption spectra of a number of substances are considered on the basis of the above statement and shown to be in accord with it. Deductions from the Valency Theory IV. Intramolecular Oscillations. J. STARK (Zeit~ch. physikal. Chem. 191 3 86 51-60. Compare preceding abstract).-A theoretical paper which con- siders intramolecular vibrations on the basis of ths Stark valency hypothesis and Planck’s radiation theory. It is shown that in an aggregate of a large number of molecules only very few valency electrons oscillate about their position of rest and that the larger number of the valency electrons have no vibration of their own apart from that of the atoms to which they are attached.At moderate temperature the valency electrons can only be made to oscillate of themselves by the absorption of light energy of the same frequency as the oscillation. If these oscillations are not damped by intramolecular collisions they take on a large ampli- tude which makes itself visible in fluorescence. It is shown that in an aggregate of molecules a portion of the atoms possess a marked thermal vibrational energy a t ordinary temperatures. The results of these deductions are applied to various chemical processes such as tautamerism. E. TALBOT PARIS and ALFRED W. PORT~R (Phil. Mag. 1914 [vi] 27 91-95),- A comparison has been made between the rotations of the plane of polarisation produced by a long column of a solution of sucrose when the light was emitted (1) from a constant source (2) from an intermittent source in the form of a platinum-mercury spark.Special’ experiments showed that the duration of the spark was less than 3 x 10-6 second. The means of several observations show that the rotations differ by less than one minute in 500° which is within the limit of experimental error. The result shows that the incident light has no directive action on the molecules of sucrose in the solution. H. M. D. J. F. S. J. F. S. Experiment on Rotatory Polarisation in Liquids. Non-influence of Oxygen on Certain Photochemical Re- actions. MARCEL BOLL and VICTOR HENRI (Cornpt. rend. 1914.158 32-34).-A criticism of Bodenstein’s theory (compare A.1913 ii 1039) explaining the disagreement of certain phot~chemical reactions with Einstein’s law of photochemical equivalence (compare J. Phys. 1913 3 277). The authors show that this theory does not apply t o the hydrolysis of the chloroplatinic acids (compare A. 1913 ii 745) or the decomposition of oxalic acid in the presence of uranyl nitrate (compare A. 1913 ii 652) these two reactions not being retarded by oxygen. W. G.GENERAL AND PHYSICAL CHEMISTRY. ii. 87 The 7-Rays of Polonium Radium and Radioactinium. A. S. RUSSELL and J. CHADWICK (Phil. Mag. 1914 [vi] 27 112-125).- Ionium radio-thorium polonium radium and radioactinium expel in addition to a-rays it minute quantity of y-radiation and this is probably generally true of all a-ray-giving substances.For these y-rays to be detected the quantity of substance required is such that it must give off as many a-particles per second as 1 mg. of radium. Ionium and polonium emit a still smaller amount of &radiation also of intensity about 10% of that of the y-radiation. A detailed account is given of the separation of polonium radium-D and -E from old radium in a form suitable for experi- ment. I n the final stage the active material is volatilised either in air or hydrogen in a porcelain tube on to a surface kept cool by contact with an inner tube in which water circulates. Polonium is volstilised in air a t 900° but to volatilise radium-E a tempera- ture above l l O O o is required although in hydrogen 900° suffices. For polonium the y-radiation emitted is of one type with p / d =215(cm.)-I which comes from the polonium itself indepen- dently of the material on which it is supported.But on copper although not on aluminium platinum or lead a characteristic radiation is excited in addition which is probably in the L series. From radium the intensity of the y-rays measured through I mm. of aluminium was from 1 to 1.5% of that of the radium in equili- brium with its short-lived products. In addition to the known soft 8-rays some hard &rays with p(Al,cm.-l) = 20 were observed which were probably due to the radium itself and not to its products. Their intensity was some 4% of that of the soft P-rays and 2.5 times that of the y-rays. I n the latter three types were recognised with ,u/d(cm.)-l in aluminium 130 6 and 0.1 respec- tively.For radioactinium two types of y-rays with p l d about 8 and 0.1 were found. This y-radiation contributes about 25% of the total y-radistion of actinium in equilibrium with its products. Results given in a previous preliminary communication on a new product in the actinium series are withdrawn. F. S. &Rays from Gases. NORMAN CAMPBELL (Phil. Mug 1914 [vi] 27 83-91) .-The theory is considered of conduction through an ionised gas when the pressure is so low that the free path of the electrons liberated is not small compared with the distance between the electrodes and the number of ions made in the gas is not large compared with the number of &rays liberated from the electrodes. The view taken is based on the work of Franck and Hertz (A. 1313 ii 548) that an electron colliding with a molecule of helium or hydrogen is reflected with little loss of energy and that with a molecule of oxygen or other element with strong affinity for the electron probably the electron is not reflected but adheres to the molecule.Some earlier measurements in air and hydrogen are considered in light of the theory and although not capable of quantitative explanation they are in general accord with it and show that the &rays from ionised gams are probably emitted withii. 88 ABSTRACTS OF CHEMICAL PBPERS. velocities of the same order of magnitude as those emitted from metals. F. S. rhe Ionisation in Various Compounds of Carbon Hydrogen and 0~ygei1 ~y kiomogaeous A-Radiation. H. MOORE (Phil. J4fig. 1914 [.I] 27 177-r87).-A long c~ Ilndrical ctrauiber of aluminium lined lnternaliy with paper to minimise the ionisation (chiefly corpuscular) due t o the radiation from the walls was filled with a mixture of hydrogen and the vapour of various liquids -inethyl formate ethyl ether ethyl formate methyl acetate benzene ethyl acetate ethyl alcohol-at their saturation pressure at Oo.The secondary X-radiation from copper was employed as the ionising radiation. It was found impossible to obtain any “ atomic ionisation ” values which would satisfy the results ; the values for an atom of carbon and for an atom of oxygen (that for hydrogen being negligible) obtained from one pair of compounds gave errors as great as 25% when applied t o the others. It was €ound that the discrepancy is due t o the different ionisation produced by a corpuscle liberated from the molecule or atom of the gas or vapour when absorbed by different gases.Further experiments showed that the real “ atomic ” effect is the liberation of these corpuscles from the atoms of the X-rays. The absorption of the X-rays resulting in the liberation of corpuscles is atomic and is the same for a given atom whether in combination or not. The ionisation produced however is a secondary effect the ionisa- tion resulting by the absorption of the liberated corpuscular radia- tion in the gas being dependent on the chemica.1 nature of the gas. Knowing the ‘ I corpuscular factor,” the ionisation in the various gases and vapours could be calculated from the composition of the gas from fixed values for the atomic corpuscular radiations of oxygen and carbon in most cases with good agreement with the experimental results F.S. Registration of the Emanation Content of the Underground Air a t Potsdam by the Benndorf Electrometer. K. KAHLER (Phyysikal. ZeitacJt. 1914 15 27-31).-A zinc cylinder 1.5 metres deep and 0.45 metre in diameter open a t the bottom was sunk into the earth and an inner electrode connected with the recording electrometer. A continuous register of the leak due to emanation in the cylinder was taken from late summer 1910 t o early summer 1912. It was found t h a t the vertical movements of the air includ- ing the changes of barometric pressure had the greatest effect on the emanation content of the air and on the escape of emanation from the ground.Sunshine also favoured the’ escape of emanation. The daily variations have no influence on atmospheric conductivity but possibly the yearly variation has an influence. F. S. A Determination of the Variation with Altitude of the Radium Emanation Content of the Atmosphere. J. R. WRIGHT and 0. F. SMITH (PhysikaZ. Zeitsch. 1914 15 31-39).-These measurements were carried out by the cocoanut charcoal methodGENERAL AND PHYSICAL CHEMISTRY. ii. 89 in Manila for two places one 5 metres and the other 2460 metres above sea-level on Mount Pauai. The individual measurements of the emanation content varied widely in accordance with the changes in meteorological conditions. Rainy weather and high wind velocities lowered whilst settled weather with low wind veloci- ties favoured the emanation content. For the low station the mean content was 82.5 x 10-12 curie per cubic metre about four times greater than the mean content 19.2 x 10-12 curie per cubic metre on the mountain.An Alleged Separation of the Radium-D from the Lead in Active Lead by means of Grignard's Reaction. CEARLES STAEHLING (Compt. rend. I 9 13 15 7 14 :30-1432).-The a u t bor has repeated the work of Hofmann and Wolfl (compare A. 1907 ii 5'2l) and has been unable to obtain any evidence in support of their views his results being entirely negative. FREDERICK SODDY (Phil. Jfug. 1914 [vi] 27 215-221).-The effects observed by Antonov (A. 1911 ii 844; 1914 ii 17) of an abnormal decay of the soft P-rays of uranium-X have been obtained from all uranium-X preparations however chemically separated in which the time of re-accumulation of the uranium-X from the uranium was short.The barium sul- phate method of separation which Antonov found gave uranium-X free from uranium-P and decaying normally gave decay curves identical with those of the other preparations. The conclusion is formed that i f uranium-P has an existence separate from that of uranium-X it must be isotop'c with it that is occupies the same place in the periodic table and is chemically identical with it. This requires that both uranium-X and -P should be formed from either uranium-Z or -ZI in a dual change in which in both modes a-rays are expelled and that initially the slope of the soft &ray recovery curve of pure uranium preparations must be considerably steeper than that due to uranium-X alone.Under conditions when the a-rays of uranium are completely absorbed the soft P-rays produce only one-tenth of the effect of the hard &rays and the point is difficult to establish. Evidence of the increased slope in the recovery curve was obtained however which although slight witis of the right order of magnitude. The whole of the results are in agreement with the view that uranium-P is the parent of actinium in a branch series starting from either uranium-Z or uranium-ZZ claiming some 8% of the atoms on the assumption that the atoms of uranium-X and uranium-P produce equal ionisa- tions on disintegration. Since however the product of uranium-P in group IV by a &ray change must be in group V a further a-ray change must occur before actinium in group 111 can be produced. Attempts to detect an a-radiation from uranium-X and -P separ- ated as rapidly as possible from uranium were negative.Dependence of the Photo-electricity of Metals on the [Sur- rounding] Gas. GEORG PAECH (Ann. Physik 1914 [iv] 43,135-164). -The dependence of the photo-electric activity of platinum F. S. W. G. The Existence of Uranium-Y. F. S.ii. 90 ABSTRACTS OF CHEMICAL PAPERS. on the nature of the surrounding gaseous medium has been examined in a similar manner to that which was adopted by Ullmann in a previous series of experiments on the photo-electric behaviour of zinc ( A m . Physik 1910 [iv] 32 1 4 8 ) . I n both cases the surrounding gas has a considerable influence on the photo- electric activity and from the data obtained with hydrogen air carbon dioxide and water vap.our i t is evident that the activity of the two metals is similarly affected by a given change in the surrounding gas.From observations in which small quantities of the vapours of methyl alcohol acetone chloroform benzene and water were admixed with air it has been found that the photo-electric effect is much larger than in pure dry air and that the difference increases with the dielectric constant of the admixed vapour. Other experiments were made in which the metal was in contact with pure ammonia methyl ether and ethylene vapours. With ammonia the photo-electric effect was five times with methyl ether 2.5 times and with ethylene 1.8 times as great as with dry air.These results also support the view that the photo-electric effect is influenced by the nature of the gas which is in contact with the active metal. H. M. D. Relation between the Conductance and the Viscosity of Electrolytic Solutions and its Bearing on the Theory of these Solutions. CHARLES A. KRAUS (J. Arner. Chem. Soc. 1914 36 35-65).-A study has been made of the conditions under which corrections for viscosity may be applied to conductivity values and the results are interpreted in the light of the ionic theory. The influence of a change in viscosity on the speed of an ion in a solution varies with the nature of the process by means of which the viscosity change is brought about. I n general such a change is effected by (1) a change in concentration due either to the addition of a new substance or of a further amount of one already present in solution; (2) a change in pressure; o r (3) a change in temperature.Consideration is given t.o the relation between the change of ionic velocity and the change of viscosity as brought about in each of these three different ways and the results obtained are based on the figures recorded by various workers in this field. It is shown that in t h O case of aqueous solutions exhibiting negative viscosity the conductivity may be corrected for viscosity in direct proportion to the fluidity change and that the same is probably true for viscosity change due t o external pressure. Corrections f o r viscosity cannot as a rule be made in the case of aqueous solutions exhibiting a positive viscosity although some- times correction may perhaps be effected by considering the influ- ence of viscosity on the individual ions.The change in the true transference numbers of electrolytes with concentration is due t o the different effect of viscosity on the speed of the two ions. The viscosity change due t o change in concentra- tion in non-aqueous solutions is positive and much greater than i t is in water for inorganic electrolytee. The velocity of the ions inii. 91 GENERAL AND PHYSICAL CHEMISTRY non-aqueous solutions changes much less than the fluidity of the solutions although at high concentrations a marked influence results. Correction for viscosity is not possible in these caws and up to N-concentrations is in many citses unnecessary. It is pointed out that' the influence of viscosity on the speed of the ions depends on the relative dimensions of the ions and molecules to which viscosity change is due.The characteristic relations observed in aqueous solutions are partly due to the complexity of the liquid water molecules. A t higher temperatures the velocities of different ions in aqueous solutions tend to reach a common limit indicating that the ions are becoming of approximately the same size. The dimensional changes observed cannot be accounted for on the assumption that the larger ions become smaller at higher temperatures but it is suggested that ~ J S the temperature rises the smaller ions increase in size owing to increasing hydration. It is shown that the view that the high speed of the hydrogen and hydroxyl ions is related to the fact that these ions are ions of the solvent itself is erroneous.Electrical Conductivity in Hydrocarbons their Halogenated Derivatives Esters and Bases. Part I. (cont.). P. WALDEN (Bull. Acad. Sci. St. PStersbourg 1913 987-996).-Further conductivity measurements have been made (compare this vol. ii 21) of solu- tions of various' substituted ammonium salts in aniline quinoline methylaniline acetic acid sulphuryl chloride ethyl f ormate methyl acetate and methyl benzoate. Consideration of the results obtained and of those previously described (Zoc. cit.) leads to the following conclusions. Beginning with benzene to1 uene and carbon tetrachloride which have dielectric constants of 2-2-2.3 all solvents are capable of giving salt solutions with measurable electrical conductivity and must be regarded as ionising media.The magnitude of the molecular conductivity depends essentially on (1) the nature of the electrolyte binary salts especially giving conducting solutions and (2) the concentration of the electrolyte; which with faintly ionising solvents must be relatively great in order that the conduc- tivity may be appreciable. In correspondence with the slight ionising power and extremely low dielectric constants of these media the ionic concentration and the molecular conductivity are only small. I n passing from the more concentrated (V equal to or less than 1) to the more dilute of these salt solutions the conductivity curve follows one of three courses according to the nature of the solvent (1) the molar conductivity A first rism to a maximum generally a t V=1-2 and t.hen falls rapidly no minimum having been yet observed; (2) A attains a maximum and then a minimum and subsequently increases again ; the dilution corresponding with the minimal or transition point varies with different solvents and the value of the minimum ranges from about 30 to 500; (3) A shows continual increase.With tetraisoamyl- or tehrapropyl-ammonium iodide case (1) is realised with solvents having dielectric constants of E. G.ii. 92 ABSTRACTS OF CHEMICAL PAPERS. about 2 case (2) when E lies between 4.95 (chloroform) and 8-10 (dichloromethane ethyl bromide quinoline s-dichloroethane) and (3) when E is greater than 9 or 10. With another binary salt having a lower ionising tendency than the above iodides for instance tripropylammonium chloride the typical curve (2) may be observed when the value of E for the solvent is greater than 9.Curve (Z) showing successively an increase a decrease and an increase of tlhs conductivity represents the general or typical form. The lengths of its first and second branches depend on the choice of solvent and solute. I n some circumstances the minimum may approach the maximum so closely that branch 2 virtually disappears and branch 3 becomes a continuation of branch 1,so that the curve exhibits no apparent maximum or minimum. Comparison of the values of A (V=20 t=25O) for tetraisoamyl- ammoaium iodide in the different solvents with the dielectric constants and viscosities of the latter shows that in general (1) in solvents with approximately equal values of 7 the greatest molecular conductivity corresponds with the greatest dielectric constant ; (2) in solvents with approximately equal dielectric con- stants the molecular conductivity is the greater the smaller the viscosity of the solvent; (3) the molecular conductivity is there- fore a function of the dielectric constant E and of the fluidity f(l!v) that is h = F ( ~ f ) .The value of the minimum shows similar dependence. Solutions of good salt-like electrolytes in solvents of high ionising power obey the Nernst-Thomson law according to which the extent of the electrolytic dissociation increases with the dielectric constant of the solvent ; further the molecular conductivity increases with the fluidity.Hence with such solvents the dependence of the molecular conductivity on the viscosity and the dielectric constant is similar to that holding with those of low ionising properties. Between the latter which are often termed insulators and solvents which readily ionise electrolytes there is then no fundamental difference. T. H. P. Some Solvents with Small Dielectric Constants. A . SACH- ANOV and J. PRSCHEBOROVSKI (Zeitsch. Elektrochew. 1914. 20 39-41). -It was previously shown (Sachanov A. 1912 ii 730) that solvents with small dielectric constants bring about a decreased molecular conductivity of the dissolved substance on dilution. With the object of testing this point further the conductivity of a number of salts in solutions of o-toluidine (DC 6-0) benzylamine (DC 5*2) chloroform (DC 4*7) bromal (DC 7*5) a-bromobutyric acid DC 7.2) and isobutyric acid (DC 2.6) were determined The speci B c conductivities of these solvents are less than 1 x 10-7.The molecular conductivities of solutions in these solvents decrease on dilution and it is shown that the relationship AvnBlh =constant holds for all concentrations; in the case of dilute solutions where H = h (approx.) the expression becomes hvn= const. where A is the molecular conductivity v the dilution and n a constant H the viscosity of the solution and h that of the solvent. I n the casesGENERAL AND PHYSICAL CHEMTSTRT. 93 examined the value found for this expression is very constant. The authors' expression is compared with a similar expression of Kraus and Bray (A.1913 ii 914) which is shown to reduce t o the same form. J. F. S. Calculation of the Electrical Conductivity in Very Dilute Aqueous Solutions. SVANTE AHRHENIUS (Medd. K. Vet. Nobelinstitwt 1913 2 No. 42 l-l2).-The author discusses the question of the correction which must be applied to the conductivity values for dilute aqueous solutions in order to eliminate the influence of the solvent. On the assumption that the conductivity of the solvent is due to carbonic acid it is shown that the methods usually adopted in the case of acids and salts are theoretically justified provided that the acids are stronger than acetic acid and that the salts in question are salts of strong acids. On the other hand if the salts ar4 salts of weak acids the correction to be applied is less than that represented by the conductivity of the solvent.The theoretical correction depends as a first approximation on the value of k / c where k is the ionisation-coefficient of the acid and c is the concentration of the salt and a table is given in which the correction is expressed in terms of the actual conductivity of the solvent such numbers being recorded for a series of values of k / c . By reference t o the conductivity measuremenh of Kohlrausch and Maltby for very dilute solutions of sodium chloride and sodium nitrate it is shown that the variation of the conductivity a t the highest dilutions is in agreement with the requirements of the mass law. F o r both salts the ionisation-coefficient k = u2/ (1 - u)u is equal to 0.024 and the view is expressed that a t sufficiently great dilutions the ionisation of all electrolytes is in accord with the law of mass action.H. M. D. Amnity Measurements in Aqueous Alcoholic Solutions. H. EULER and I(. BLOMDAHL ( A ~ k i v . Kern. Min. Geol. 1913 4 No. 40 l-8).-The authors have determined the affinity constants of phenol and guaiacol in 80.4 aqueous-alcoholic solution by measure- ments of the conductivities of solutions of their ammonium salts (compare Euler and Bolin A 1909 ii 374; Hagglund A. 1912 ii 120 910). The values are respectively 3-1 x 10-12 and 2.4 x 10-12 corresponding with the fact that the dihydroxybenzenm are very little more dissociated than phenol (Euler and Bolin The ratio of tho dissociation constants of the compounds salicylic acid acetic acid pnitrophenol and phenol in water to that in 80% alcohol varies between 15 and 77; experiments with phosphoric acid indicated that its dissociation is affected in a similar way probably to a greater extent by alcohol.loc. cit.). T. S. P. Dissociation Constants of Organic Acids. JAMES KENDALL (Medd. K. Vet. Nobelinrrtitzct 1913 2 No. 38 1-27).-A general formula for the ionisation of acids has been given in a previous paper (T. 1912 101 1275) and further measurements of theii. 94 ABSTRACTS OF CHEMICAL PAPERS. electrical conductivity of acids have been made in order to test the formula more completely. The acids previously examined have been investigated in more concentrated solutions and measurements have also been made for salicylic acid and some of its derivatives.The results obtained afford further confirmation of the view that the ionisation constant is subjected to the influence of two disturb- ing factors clne of which (the total concentration influence) causes the constant to diminish as the concentration increases whilst the other (thc ionic concentration influence) gives rise to an increase in the constant with increasing concentration. I n the case of very dilute solutions of salicylic acid and its derivatives it is found that the ionisation constant increases to a small extent with increasing dilution. This property is character- istic of dibasic acids and it is supposed that the proximity of the carboxyl group to the phenolic group increases the strength of the latter to such an extent that the acid behaves as a weak dibasic a,cid.I n agree'ment with this it is found that salicylic acid and its derivatives cannot be titrated a t all sharply with phenolphthalein as an indicator. m2/ (1 - m)q= L + c(1- m ) / m is not apparent but the presence of the expression (1 - m ) as a factor on the right-hand side of the equation seems to point t o the activity of the non-ionised acid aB being the cause of the variation in the constant. It is also shown that the author's equation may be regarded as a special form of the general equation put forward by Bray and Krczus (A 1913 ii 914). The theoretical significance of the equation H. M. D. Piezochernical Studies. XII. The Influence of Pressure 03 ABnity. 111. G. TLMOFEEV (Zeitsch. physikal. Chem. 1913 86 113-128. Compare A.1905 ii 678; 1911 ii 15).-The influence of pressure on the E.M.K. of the elements (a) Ag-AgC1-KC1 0-1N-HgCl-Hg ; ( b ) 7% Zn amalgam-35-037L ZnClz solution-AgC1-Ag; ( c ) 7% Zn amalgam-65*59% ZnC1 solution-AgCl-Ag ; (d) 7% Zn amal- gam-65-59 ZnC1 solution-HgC1-Hg has been determined over a range of pressure 1-1500 atmospheres and a t 25O. The influence of pressure can be represented by the expression in which E and E represent the E.M.F. at pressures 7r and 1 atmosphere respectively and v1 and v2 the volumes of reacting substances before and after the change. E - E ~ = T ( v ~ - v ~ ) J. F. S. Calibration Tables for Copper-constantan and Platinum- platinrhodium Thermoelements. L. H. ADAMS (J. Amer. Chern. Xoc. 1914 36 65-72).-0wing t o the fact that thermo-elements are becoming more extensively used for measuring high temperatures two tables have been conetructed with the object of obviating trouble in converting micro-volts into degrees.Each table gives the temperature and temperature-difference for every 100 micro-GENERAL AND PHYSICAL CHEMISTRP. ii. 95 volts the cold junction being maintained a t Oo. The table for each of the two elements is intended for use in conjunction with a deviation curve determined by calibration of the particular element with three or more of certain fixed points given at the foot of each table. E. G. Behaviour of Wehnelt Electrodes in Different Gases. KARL FREDENHAGEN (Physikal. Zeitsclh. 1914 15 19-27. Compare Gehrts A. 1913 ii 1005).-The influence of oxygen nitrogen and hydrogen on the efficiency of the Wehnelt cathode has been investigated by measurements a t 9 4 5 O 1085O and 1235O and in a few cases a t 1340O. The results obtained are presented in the form of curves which show the connexion between the current and the applied potential up to 200 volts.In all cases the current-potential curves were determined for different pressures of the gas present. These curves indicate that nitrogen diminishes the efficiency of the oxide electrode but has little influence on the variation of the effect with temperature. I n presence of oxygen the efficiency of the electrode is found to be diminished t o a very much greater extent and at the same time the temperaturecoefficient of the effect is reduced especially a t the higher oxygen pressures. I n contrast with the behaviour of nitrogen and oxygen hydrogen is found to increase the activity of the electrode and in this case the currentcpotential curves for the different temperatures lie very close together.A few experiments have also been made with methane coal-gas and cyanogen. The two former behave like hydrogen whilst cyanogeu reduces the activity of the electrode to a much greater extent than oxygen. The change produced by the cyanogen in the oxide electrode is permanent and the activity is not restored even when the electrode is strongly heated in an atmosphere of oxygen. It is supposed that the change is due to the formation of calcium cyanamide. The observations are interpreted in terms of the theory of the Wehnelt electrode which has been put forward in a previous paper (Zoc.cit.). H. M. D. [Electrodes of the Third Type. J FRIEDRICH AUERBACH (Zeitsch. pl~yaikcd. Chem. 1914 86 243-246).-A criticism of the methods employed by Spencer (A 1913 ii 596) in the calculation of the ionic solubility product L = [Cu"] [I03/]2 for cupric iodate. It is shown that the agreement between the values obtained from different combinations of electrodes is determined by algebraic relationships. From Spencer's data (compare also A. 1912 ii 1129) the author finds L = 1.47 x 10-7. In view of the magnitude of the solubility of cupric iodate it is not permissible to calculate this solubility directly from the ionic solubility product. E. M. D. A Scheme for the Dissociation of Ternary Electrolytes. RUD. WEGSCHEIDER (Zeitsch. Elektrochent. 192 4 20 18).-Polemical against Drucker (A 1913 ii 1015).The author shows thatii. 96 ABSTRACTS OF CHEMICAL PAPERS. Drucker’s statements are too wide if he applies the generally adopted meaning to the word “electrolyte,” but if the term ‘ I electrolyte ” is distinguished from the term “ half-electrolyte ” the conclusions arrived a t by Drucker are tenable. J. F. S. The Electrolytic Deposition of Alloys and their Metallo- graphical and Mechanical Investigation. I. Alloys of Nickel and Iron Deposited at the Ordinary Temperature. R. KREMANN C. TH. SUCEY and R. MAAS (~ki’onarsh. 1913 34 175‘i-lS~~9).-All the experiments were made using a current density of one ampere per sq. dcm. a t the cathode. Iron .may be deposited from solutions of a ferric salt (0-5-M ferric ammonium sulphate) the current yield being 3*9% if it is assumed that the Fe”’-ions are directly deposited.If nickel sulphate (0.6-Jf solution) is also present in solution the current yield is increased agreeing with the assumption that in the simul- taneous deposition of two metals which form solid solutions the deposition of these metals is facilitated by the mutual depolarising effect of the metals. I n all the following experiments the electrolyte consisted of a solution of ferrous and nickel sulphates in varying proportions. It was found that the ratio (R,) of nickel to iron in the alloy was considerably smaller than the ratio (R2) of the metals in ths solution; also R gradually increased with R,. The character of the deposit depended on the nature of the cathode; with a carbon cathode the deposits were smooth and adherent whilst with iron cathodes they were irregular and scaly; in all cases they were brittle owing to occlusion of hydrogen.When a combined nickel- iron anode was used chiefly iron went into solution. A slight alteration of the current density had scarcely any influence on the composition of the deposit. The addition of potassium oxalate to the electrolyte increased the ratio R in a,ccordance with the fact that nickel forms more highly complex ions with oxalate-ions than does iron. Attempts to coat sheet iron with adherent deposits of an iron- nickel alloy from a neutral electrolyte were unsuccessful owing to the fornation of large quantities of a cathode sludge and to the fact that the expansions of sheet iron and the alloy deposited are different.The alloy was also very brittle because of its hydrogen content and in different experiments carried out under the same conditions the composition of the alloy varied. The formation of the cathode sludge could be prevented by the addition of sulphuric acid to the electrolyte but in solutions 0*2N with respect to the acid no deposition of metal took place. I n the last seriea of experiments citric acid was added to the electrolyte. It was found that the acid was used up during the electrolysis and a t the same time the alloy deposited contained carbon which varied from 0.5 to 1.1%; the higher the concentra- tion of the citric acid the greater the carbon content of the alloy. With increase in value of the ratio R the carbon content of the alloy increased to a maximum and then fell.When the concen-GENEBAT ASD PHYSICAT CHEMISTRT. ii 9 i tration of the citric acid was kept constant the ratio R iiicreased with t-he ratio R although not in proportion. The addition of citric acid to the electrolyte almost completely did away with the formation of cathode sludge and it was possible to deposit alloys 4-5 mm. in thickness on a carbon anode; the alloys were brittle. The metallographic investigation of the alloys deposited showed that they were' analogous in structure to and in some cases identical with alloys prepared by thermal method@. In some cases structures similar to those occurring in meteorites were observed. A structure whicn was characteristic of the electrolytic alloys consisted of concentric layerings (sphmolites).The cross-section of the deposits had a distinctly lamellar appearance which only disappeared on heating to a white heat. The hardness of the various electrolytic alloys wils measured by the sclerometric method. Both nickel and iron when deposited from acid baths are softer than when deposited from neutral baths. The alloys obtained from acid baths are harder than would corre- spond with the rule of mixtures. The authors are of the opinion that the hardnew of the electrolytic deposit depends on the fineness of the structure and not on the presence of hydrogen; the finer the structure the harder the alloy. T. S. P. Magnetic Properties of the Alkali Metals in Combination. PAUL PASCAL (Compt. rend. 1914 158 37-39).-By measuring the molecular-coefficients of magnetisation of the haloid salts and the nitrates sulphntes and carbonates of the alkali metals knowing the values of the part contributed by the acidic ions the author has calculated the atomic coefficients of magnetisation of these metals.The values obtained increase with increase in atomic weight. By plotting the logarithms of these coefficients against the atomic weights he finds that the points for lithium sodium and potass- ium lie on a straight line and for potassium rubidium and czesium on another straight line. Comparing these figures with those obtained for hydrogen and silver hydrogen comes on the first line and silver on the second line. W. G. The Shape of Molecules. T. SVEDBERG ( A ~ k i v Kern.Alin. Geob. 1913 4 No. 39 1-5).-1t has been shown by Cotton and Mouton (compare A. 1913 ii 4) that certain isotropic substances for example nitrobenzene become doubly refracting in a strong magnetic field. This is probably due to incomplete orientation of the molecules but hitherto i t has not been possible to detect any variation of the electrical conductivity of such substances according as the conductivity is measured along or across the lines of magnetic force. Such a variation should occur if the molecules show geometrical dissymmetry since the resistance to migration should vary with the orientation of the molecule. The author has found that such a variation takes place in the conductivity of anisaldazine. This substance forms an anisotropic liquid (liquid crystals) between 165O and 180° and in a magnetic VOL.CVI. ii. 7ii. 98 ABSTRACTS OF CHEMICAL PAPERS. field behaves as a strongly doubly refrating uniaxial crystal the optical axis of which is parallel with the magnetic lines of force (compare Wartenberg A.. 1911 ii 952; Mauguin A. 1912 ii 630). Since the amount of ionisation and consequently the conduc- tivity of anisaldazine is very small an electrometric method of measurement had to be used for determining the conductivity. The results showed (1) that the conductivity in a magnetic field is smaller than the ordinary conductivity; (2) that with the cell used the relative increase in resistance at right angles to the magnetic lines of force is about three times greater than parallel to the same; (3) that in a field of about 4500 Gauss the relative increase in resistance t o the magnetic lines of form is of the order of 4%; (4) that the influence of the magnetic field is closely connected with the anisotropic condition of the anisaldazine since it does not exist above 180° or below 165O.T. S. P. ETpansion of Mixtures of Garbon Disulphide and Acetone. J. VECINO Y VARONA (Anal. Pis. Quirn. 1913 11 498-515).-The density a€ Oo and refractive index a t 14'58O of acetone were found to be 0.81229 and 1.43528 and of carbon disulphids 1.29217 and 1.63149 respectively. The coefficients of expansion of acetone and carbon disulphide were determined as 148.1 x 10-6 and 117.53 x 10-5 respectively. Mixture of these liquids is attended by expansion in contra- distinction to the usual rule and similarly the thermal expansion does not follow the rule of mixtures possibly on account of the formation of molecular compounds. G.D. L. Convenient Method for Calibrating Thermometers by means of Floating Equilibrium. THEODO~ZE W. RICHARDS and JOHN W. SHIPLEY (J. Amev. Chem. Soc. 1914 36 1-lo).-In an earlier paper (A. 1912 ii 599) it was shown that the (' floating equilibrium " of a suitable solid float is exceedingly sensitive and that the " floating equilibrium temperature " is almost a linear function of the concen- trations. It was also shown that this method can be used t o analyse solutions and to calibrate thermometers. The present paper deals with the application of this process to the standardisation of thermometers and experiments are described which consist in determining the temperatures at which different aqueous solutions of sodium sulphate attain exactly the density of a given float.By plotting the floating equilibrium temperatures against the percentage composition of the solutions an almost linear curve is obtained and it is pointed out that this curve when once established will serve for the calibration of any thermometer a t any time. The method is also serviceable for determining the coefficients of expansion of liquids with known floats or the coefi- cients of expansion of floats with known liquids. E. G. Electrical Contact Vapour-pressure Thermoregulator. ALEX. L. FEII.D (J. Am#. Chem,. Soc. 1914 36 72-76) -A new form ofGENERAL AND PHYSICAT CIIEMISTRT. ii. 99 vapour-pressure thermoregulator is described which is not affected by changes of atmospheric pressure.It consists essentially of an upright U-tube closed a t both ends. One side of the tube contains a volatile liquid the b. p. of which is near the desired temperature of the bath. Above the liquid is its saturated vapour with or without admixture of air or other gas. The pressure exerted by the vapour or the vapour and gas supports a column of mercury which maken and breaks an electrical contact on the other side of the U-tube. This contact is made in an atmosphere of dry carbon dioxide. This thermoregulator is light easily adjustable and has an observed sensitiveness of 0.01-0*005°. It is adapted for use in liquid baths is not affected by agitation of the liquid and rapidly adjusts itself to changes of temperature of the bath.The regulator is therefore particularly adapted for the accurate control of the temperature of an ordinary thermostat. Speciflc Heat. 0. D. OSGOOD (Zeitsch. RZektrocIwtn. 1914 20 37).-A mathematical paper in which the application of a curve found in a thermodynamical investigation of the mass action equilibrium constant is shown in connexion with a heat loss as indicated from the magnetic hysteresis curves. Speciflc Haat of Solids. 11. H. VON JUPTNER (Zeitsch. Elektro- chsm. 1914,20 10-18. Compare A. 1913 ii 921).-A msthematicrtl paper in which an expression is evolved by means of which the atomic heat of solid substances between the temperatures T = Pv' and I1=O.394/3v' can be calculated. An expression of the form E.G. J. F. S. is deduced for the mean atomic heat of solids between T=Pd and higher temperatures. A series of tables are given by means of which the true and mean atomic heats and heat capacities between 1'=0.394Pvr and any higher temperature may be calculated. It is shown that the heat capacity a t temperatures below T = 0*394Pvr is so small that i t for most purposes may be neglected. ALBERT WIGAND (Zeitscli Elekti-ochem. 1914 20 38-39).-The specific heat of the two forms of tin have been redetermined by the author and it is shown that grey tin has a smaller value than white tin. This is in accord with the measurements of Bronsted (A. 1912 ii 897). The previous valum for grey tin obtained by Wigand (A. 1907 ii 70) are shown to be due to the presence of oxide.These results are discussed in connexion with the van't Hoff- Thomsen rules and the Richarz rule the latter of which states that in the case of allotropic substances the form with the higher specific heat has the smaller density. This rule is not followed in the case of tin. J. F. S. An Invereion Point for Liquid Carbon Dioxide in Regard to the Joule-Thomson Effect. ALFRED W. PORTER (Proc. Koy. SOC. 1914 [ A ] 89,3'77-378).-According to the experiments of Jenkin and J. F. S. Specific Heat of Grey Tin and the Richarz Rule. 7-2ii. 100 ARSTRACTS OF CHEMICAL PAPERS. Pye (A. 1913 ii 753) the Joule-Thomson effect for liquid carbon dioxide shows a change in sign a t -ZOO to -3OO. Assuming that the change in pressure may be treated as a differential it is shown that the inversion point must correspond with a minimum (or maximum) value of v/!P where t~ is the specific volume at To (absolute).Three sets of experimental data corresponding with a mean pressure of 500 .pounds per ,square inch concur in giving a minimum value of v / T a t about -24O which agrees fairly closely with the result of direct observation. The result is of special interest in that it implies that liquid carbon dioxide in this region behaves approximately like a perfect gas its volume being nearly proportional t o the absolute temperature. Latent Heat of Fusion of Ice. H. C. DICKINSON D. R. HARPER and N. 5. OSBORNE (Chem. Zentr. 1913 ii 2025 ; from J. Franklin Inst. 1913 176 453-454).-The latent heat of fnsion of ice has been determined (I) by allowing a mass of 100-200 grams to melt in a calorimeter and (2) by measurement of the energy in form of an electric current necessary to fuse a mass of about 500 grams placed in a calorimeter.As mean of ninety-two experiments performed with a great variety of samples of ice the number 79.63 cal. per gram is found. Occurrence of an Upper Critical Point of Mixing at the Go-existence of Two Mixed Crystal Phases. A. S m ~ s a n d C. A. LOBRY DIG BRUYN (Proc. K. Akud. Wetensch. Arnstwdum 1913 16 557-564).-The freezing-point and melting-point curves for binary mixtures of potassium and sod;um chlorides are continuous and exhibit a minimum at about 664O. During the cooling of a solidified mixture of the two salts changes occur whereby the homogeneous isomorphous mixture is converted into two co-existing mixed crystal phases.These phases are identical in composition at 407O but as the temperahwe falls they diverge t o an increasing extent until at the ordinary temperature the one consists of practically pure potassium chloride and the other of pure sodium chloride. This mixture affords therefore an instance of an upper critical mixing point. I n order to ascertain whether the observed phenomena are due t o the formation of a compound experiments have been made with two mixtures containing respectively 20 and 60 mols. % of potassium chloride. The two mixtures were melted and allowed to cool the rate of cooling between 37OC and 340° being very slow. The mix- tures were then powdered and agitated with a little water for two hours a t 25O the saturated solutions obtained being then analysed.The analyses show that the same solution was obtained for both solid mixtures and from this the conclusion is drawn that the phe- nomena of mixing in the! solid state are not due t o the formation of a compound of the two salts. G. WYROUBOFF ( A r m . Chim. Ph,ys. 1913 [viii] 30 633-634. Compare A. 1909 ii 740).-A reply to Cornec (compare A. 1913 ii 840). H. M. D. H. W. H. M. D. The Basicity of Acids. W. G.GENERAL AND PHYSICAL CHEMISTRY. ii. 101 Properties of Liquids as Functions of the Critical Constants. JAMES KEP~DALL (Medd. K. Vet. Nobelinstitut 1913 2 No. 29 1-28).-1t is shown that the heat of vaporisation and the specific volume can be expressed as simple functions of the critical con- stants and by reference to the experimental data of Young (A. 1910 ii 271) the validity of these relations is demonstrated for the ordinary range of temperatures.For the heat of vaporisation the equation obtained is &=k(T,-T)@ in which Q is the latent heat T the temperature T the critical temperature L a constant characteristic of the liquid under exam- ination and n a constant which has the same value 0.386 for all normal liquids. By combination of this equation with Trouton’s rule i t assumes the form MQ = 20*7T(1 - 7’/ T,)n in which M is the molecular weight of the liquid. For the specific volume of liquids the equation obtained is JT- $x=kl( T - T$ in which d is the density of the liquid at TO d the critical density and k1 a constant which depends on the nature of the liquid.The variation of the density of the saturated vapour with the tempera- ture is reprwented by a similar formula in which D is the density of the saturated vapour a t temperature T and k is a constant characteristic of the substance. The densities of liquids a t absolute zero have been calculated from the first of these two equations and it is found that the values obtained in this way are in good agreement with those found by Guldberg by extrapolation. The ratio of the density a t absolute zero to that a t the critical temperature is approximately the same for all normal liquids with an average value of 3.75 which is identical with the value obtained by Guldberg. With the aid of this relation d,!d = 3.75 the equation connecting the specific volume of the liquid with the temperature may be simplified in which case it reduces to the form $/d/d,-1=0*553(1 -T/T,)+ and this equation is found to be approximately satisfied by all normal liquids.By combination of the above equations with that of Mills for the heat of vaporisation the relation is obtained in which P is the vapour pressure V the volume of 1 gram of saturated vapour ’u that of 1 gram of liquid and p is a constant charactenstic of the liquid. According to this the equation Q = k( T - T)O.s*6 is an interpolation formula in which the exponent has a value intermediate to the values which are charac- teristic of the second and third terms of the previous equation. At low temperatures this equation assumes the form I n this equation the condition d&,/dT = 0 is satisfied if the relation- ship kipM/3T,t=R holds.This is found to be the case for normal liquids. H. M. D. yx - yo= kL( 7; - T,& Q =O*O431833P(V -v) + k,p(T,- T)*+ kZp(Tc- T)+ Q = R T I M + k ( T - T ) 8 + f/. J d,.ii. 102 ABSTRACTS OF CHEMICAL PAPERS. Heat of Vaporisation of Associated Liquids. JAMES KENDALL (Medd. K. Vet. Nobezinstitut 1913 2 No. 36 l-l4).-The formula for the heat of vaporisation of normal liquids given in a previous paper (preceding abstract) has been applied to a number of associated liquids. I n the case of normal liquids the value of n in the equation &,=k(T,-T)rh is equal to 0.386 but it is found that the exponent must have its value changed if a formula of this type is to afford values in agreement with those derived from experi- ment.I f the heats of vaporisation of water acetic acid and methyl ethyl and propyl alcohols are plotted against the temperature (TC-L") it is found that water approximates most nearly to the normal type whilst acetic acid is the most abnormal. The alcohols occupy an intermediate position. An explanation of the difference in behaviour is put forward in which the changes in molecular complexity on vaporisation and the accompanying heat changes are taken into consideration. H. M. D. New Relations between the Physical Properties of Liquids. DANIEL TYREH (J. Physical Chem. 1913 17 717-736).-A large number of empirical relations have been found connecting the latent heat of vaporisation surface tension density and other physical properties of liquid substances. From an examination of these relations by reference to the data for a number of chemically dissimilar substances it appe,ars that the three following equations are satisfied within the limits of experimental error.I n these equations I is the internal heat of vaporisation dl the density of the liquid d that of th4 saturated vapour T the absolute temperature T the critical temperature T the boiling point d8 the density of the liquid a t T y the surface tension m the molecular weight C (71.. . constants which are independent of temperature but vary with the nature of the liquid k k f . . . constants which are inde- pendent both of the temperature and the nature of the liquid. The three relations in question are y+=Vdl+(d$- d,3) where C'=kT,+/m+. d,a) and ydsa /Z(& -&)? = k'.(TG- T)dt/Z(dl-d,))- =Cff=kffm/d, holds equally well provided the temperature is not less than 6OC below the critical temperature and the relation ml= kf f f J T,(T,-T) has a similar validity if the temperature is more than 80° removed from the critical temperature. The other relations given by the author are approximate in character and the divergences met with when these are applied to actual data cannot be accounted for on the basis of experimental error. H. M. D. Relation between the Boiling Point and Composition of a Mixture of Ethyl Iodide and Ethyl Alcohol. SARAT CHANDRA JANA and JITENDRA NATH SEN GUPTA (J. Amel.. Chcm. ACoc. 1914. 36 115-118).-A study has been made of the b. p.'s of mixtures of ethyl iodide and ethyl alcohol and it has been found that these are much lower than the b.p. of either of the two components. A ml = C( T - T)+/ (d+ + d$) The equation :GENERAL AND PHYSICAL CHEMISTRY. ii. 103 mixture containing 26.69% of ethyl iodide has b. p. 7 0 0 2 ~ . As the proportion of ethyl iodide is increased the b. p. gradually falls until a minimum is reached with a mixture consisting of 87% of ethyl iodide and 13% of ethyl alcohol. This mixture distils a t 61-2O as a homogeneous liquid. On further increasing the proportion of ethyl iodide the b. p. rises again; when 97.63% of ethyl iodide is present the b. p. is 64*0°. The results of the investigation are plotted as a curve. General Cornparkon of Vapour Pressures. J. M. CRAFTS (Compt. rend. 1913 157 1403-1405).-By means of the formula T - TI = (TI1 - T/ll)C where T and TI are the boiling points of any substance at pressures P and PI and T N and TI1’ are the boiling points of a standard substance a t the same pressures C being a constant it is pomible in most cases to calculate all other vapour pressures of a substance.I n the case of hydrogen chlorine water alcohols and acids the formula needs to be modified t o where c is another constant. The standard substance used is naphthalene and a table is given showing the results for a considerable number of substances and giving the values of C and c. The value of t l - t increases by unity from zero for each 20° difference between TI1 and TI11 for nap h t-h alene. Cheaard’s Tube for Fractionation by Adiabatic Condensa- tion. E. HILDT (BUZZ.SOC. chim. 1914 fiv] 15 37-41).-The E. C. T - TI = (TI1 - T”’)[C ~ ( $ 1 - t ) ] W. G. apparatus is shown in the diagram. Partial condensation takes place in the spiral S owing according to the author to the cooling of the vapour brought about by the work done in expand- ing against the external pressure. The liquid thus condensed isii. 1104 ABSTRACTS OF CHEMICAL PAPERS. returned by the tubes 0 Of 0'1 etc. to the tube L and hence to the flask. Using this apparatus the first four fractions of 40 C.C. obtained in the distillation of 1500 C.C. of a 12*50/ alcohol con&sted of 91% alcohol. T. S. P. The Temperature of Sublimation. J. JOLY (Phil. M G ~ . 1914 [vi] 27 1-14. Compare A. 1913 ii 556).-Further experiments have been made to determine the approximate temperature a t which sublimation occurs when different minerals are heated t o a gradually increasing temperature in contact with air.The new obsmvations have reference t o minerals containing antimony arsenic sulphur tellurium bismuth selenium and lead and it is shown that the temperature of sublimation may be utilised in the classificatiori of the members of these different groups. H. M. D. Laboratory Apparatus for the Continuous Evaporation of Large Volumes of Liquid in a Vacuum. WILLIAM A. DAVIS ( J . Ayric. Xci. 1913 5 434-436).-8 simple appamtus is described with sketch by means of which large volumes of alcoholic plant extracts which are liable to froth can be evaporated continuously and the distillate recovered if desired in fractions. When once started the distillation can be left to itself except when the liquid in the flask has to be renewed.The apparatus consists of an ordinary distilling flask the side- tube of which is bent upwards and passes into a short wide tube which serves as a froth trap. The latter is connected with the top of an upright condenser the lower end of which is fitted to a cylindrical funnel (with tap) which is connected below with a pump flask and above with a large reservoir t o ensure a steady vacuum. The vacuum is obtained by means of an ordinary water- pump a safety valve (Hutchinson A. 1911 ii 933) being inserted between the pump and the apparatus. N. H. J. M. Heat of Formation of Hydrogen from Hydrogen Atoms. IRVING LANGMUIR (PhiE. Mag. 1914 [vi]. 27 188-189. Compfire A.1912 ii 826).-Further measurements of the loss of heat from electrically-heated tungsten wires in hydrogen a t pressures ranging from 1 t o 760 mm. have shown that the value of 130,000 cal. for the heat of formation of hydrogen molecules from hydrogen atoms is much too large. The new experimental evidence indicates a value of 75,000 t o 80,000 caI. which is much closer t o the value calculated by Bohr from his theory of the constitution of atoms and molecules (A. 1913 ii 689 943 1045). H. M. D. The Determination of Heat of Formation from Electro- motive Force [Measurementsl. LUISE WOLFF (Zeilsch. E/&trochem. 1914 20 19-22).-The electromotive force and temperature- coefficients of the E.M.F. have been meaaured for the cells (I) Ag I AgCl 1 nHCl1 C1,1 Pt (2) Hgl HgCII nHC11 C1 I Pt (3) Ag.] AgC'l I nHCl I HgCl1 Hg,GENERAL AND PHYSICAL CHEMISTRY.ii. 105 and from the values obtained the heat of formation of silver chloride and mercurous chloride and the heat of reaction of the change 2Ag + Hg2C1 = 2Hg -+ 2AgC1 have been calculated. The chlorine electrode was fed by chlorine largely mixed with air to reduce the secondary reactions. The heat of formation of silver chloride is found to be 30612 cal. Hg,C1 31827 cal. and the heat of the reaction between silver and calomel -2.1324 cal. A com- parative table is given of the corresponding values found by Halla Thomsen Bronsted Korev and Braune. J. F. S. The Heat of Formation of Some Compounds of Cupric Chloride with Ammonium Chloride. A. BOUZAT and ED. CHAUVENET (Compt.rend. 1914 158 40-42).-The authors have determined the hects of solution of anhydrous cupric ammonium chloride CuC1,,2NH4C1 and the hydrate with 2H20 obtained in the blue or the green form. The' anhydrous double chloride is best prepared by warming the hydrate in an atmosphere of hydrogen chloride and chlorine or heating it at looo in a current of dry chlorine. W. G. Thermochemical Studies of Diazo- and Azo-compounds. VI. Diazo-derivatives of t h e Three Nitroanilines. V. SVENTO- SLAVSKI (J. Buss. Phgs. Chsrn. Soc. 2913 45 1789-1765. Compare A. 1910 ii 691).-The heat of diazotisation of m-nitroaniline NO,*C,H,*NH,(diss.) + HCl(diss.) + HNO,(diss.) = 2H20(liq.) + NO,*C,H,*N,Cl( diss.) in aqueous hydrochloric acid is +24-19 Cal. and in acetic acid + 16.75 Cal.The heat of the reaction N0,*c6H4*N2C'l(diss.) + 2NaOH(diss.) = N0,*C6H,*N,-ONa(diss.) + NaCl(diss.) + H,O(liq.) has been measured in three ways (1) starting from the anti-cjalt + 11-20 Cal. ; (2) starting from the amine + 11-04 Cal. ; (3) starting from the diazonium chloride + 11-58 Cal. The agreement between these numbers in conjunction with the fact that the last method excludes the possibility of the formation of traces of symdiazo-salt demonstrates that the action of sodium hydroxide even in dilute solution on m-nitrobenzenediazonium chloride consists in the quantitative isomerisation of the m-nitrobenzenediazonium into the anti-diazo-salt. Measurements of the amounts of heat developed in the coupling of m-nitrobenzenediazonium chloride with @-naphthol show that the process proceeds quantitatively independently of whether the &naphthol or the diazonium salt is in excess.The value +18.89 Cal. previously given (A. 1910 ii 691) for the heat of diazotisation of pnitroaniline is incorrect owing to an error in the calcizlation of the amount of sodium hydroxide used in one of the control experiments. I n aqueous solution the value for the reaction N02*C,H4*NH2,HCl(diss.) + €INO,(diss.) = N02-C6H,*N,Cl(diss.) + 2H,Oii. 106 ABSTRACTS OF CHEMICAL PAPERS. is now found to be +19-722 Cal. in aqueous solution and +18*36 Cal. in acetic acid or allowing for the heat of solution of p-nitro- aniline in acetic acid namely -4.47 Cal. (Zoc. cit.) +13.89 Cal. The heat of the reaction II\To,*c,H40Nz~~(d~s.) + 2NaOH(diss.) = N02*C,H,*N,*ONa(diss.) + NaCl(diss.) + H%O(liq.) is + 13.77 Cal.or if the sodium hydroxide is taken in large excess + 14,724 Cal. so that the salt NO,*C,H,*N,*ONa undergoes hydro- lysis but to an extent inconsiderable in comparison with that shown by salts of cis-diazo-compounds. The heat of the reaction is 6.513 Cal. and that of N0,*C,H4*NzC'1(diss.) + NaOH(diss.) = ~ O * C ~ ~ N ~ ~ ~ ( s o l i d ) + NaOH(diss.) = N0,°C,H4*~2*O13[(solid) + NaCl(diss.) NO,*C,H,*N,*ONa(diss.) + H,O + 6.991 Cal.; the sum of these two values + 13'50 Cal. agrees satis- factorily with the number given above +13*77 Cal. Hence the product NO,*C,H,*N,*OH ,precipitated by the action of thO first molecule of sodium hydroxide does not undergo any subsequent profound change but these results are insufficient t o show that it5 N*OH ' structure is that of a trans-diazonium hydroxide since i t might equally well be the nitrosoarnine NO,*C,H,*NH*NO.But if it is assumed that the heat of solution of the precipitate differs little from that of the nitroaniline 3-74 Cal. the heat of the reaction NO,. C,,Hp*# NO,-C,&*N,*OH(diss.) + NaOH(diss.) = NO,*C,H,*N,-ONa( diss.) + H,O(liq.) will be 3-74 + 6-99 = 10.73 Cal. ; and the heat of the same reaction taking place under such conditions that the compound undoubtedly has the structure of a trans-diazo-hydroxide has already been determined to be +10*37 Cal. (Zoc. cit.). Comparison of the heats of solution of pnitrobenzenediazonium chloride in 12.7% hydro- chloric acid - 1.363 Cal. and in water - 3.00 Cal. indicates hydro- lysis in the aqueous solution.The tram-salt NO,*C,H,*N,*ONa after repeated crystallisation from 95% alcohol contains 2H,O and not lH,O as was stated by Schraube and Schmidt (A. 1894 i 237). Its transformation in 12.7% hydrochloric acid into diazonium salt N0,*C,H4~Nz*ONa(diss.) + 2HCl(diss.) = NO,-C,H,*N,Cl(diss.) + NaCl(diss.) + H,O(liq.) is accompanied by the development of 15-48 Cal. and from this the heat of the reaction NO,*C,R,*N,Cl(diss.) + 2NaOH(diss.) = N02*C,H4*N,*ONa(diss.) + NaCl(diss.) + H,O(liq.) is + 11.92 Cal. The difference between the latter number and that determined above +13*77 Cal. is owing to hydrolysis and decom- position of the tram-salt at the beginning of the reaction to the unevenness of the process of isomerisation and to the difficulty experienced in making use of concentrated hydrochloric acid as a calorimetric liquid; 13'77 Cal.is the more probable value andGENER.AL AND PHYSICAL CHEMISTRY. ii. 107 13.63 Cal. instead of 15.48 Gal. for the reverse change. effect of the reaction The heat NO,*C,H,-N,Cl(diss.) + P-C,,H,*ONa(diss.) = Na C1 (diss. ) + NO,* C,H,* N C,,H,* OH( solid) is +38*69 Cal. which should differ from 26.21 Cal. the heat developed by the same reaction in 95% acetic acid by approximately the heat of neutralisation of acetic acid namely 13.30 Cal.; the actual difference is 12.48 Cal. With o-nitroaniline the following results were obtained NO,-C,H,*NH,,HCl(diss.) + HNO,(diss.) = NO,-C,H,~N,Cl(diss.) + 2H,O(liq.) + 18.66 Cal. NO,*C,H,*N,Cl(diss.) + ZH,O(liq.) + 13.07 Cal.or in acetic acid N0,*C,H4-NH,(solid) + HNO,(diss.) + HCl(dies.) = For the isomerisation of the diazonium salt into tram-diazo-salt by the action of excess of sodium hydroxide the amount of heat developed +13*78 Cal. agrees exactly with that found with the corresponding para-derivative. These resulk in conjunction with some of those previously obtained show that for aniline p-toluidine a-naphthylamine and m- and pnitroanilines the sum of the heats of diazotisation isomer- isation into diazo-hydroxide and coupling is constant 48.44-49.35 Cal. For o-toluidine benzidine and dianisidine the heat of neutralisation by hydrochloric acid necessary to the calculation is unknown. T. H. P. Thermochemical Studies of Diazo- and Azo-compounds. VII. Supplementary Data for Diazobenzenesulphonic Acid.V. SVENTOSLAVSRI and A. MANOSZON ( J . Rugs. Phys. Chem. hbc. 1913 45 1765-1770).-The following reactions have been investigated (1) Using excess of the nitrite S03H*C,&*NH2(diss.) + NaNO,(diss.) = SO,Na~C,H,=N,=OH(diss.) + H,O(liq.) + 16.20 Cal. Using excess of the sulphanilic acid rnSO,H*C,H,*NH,(diss.) + NaNO,(diss.) = SO,Na*C,R,*NH,(diss.) + c,H,i<r (diss.) + 2H,O + so (m - 2)SO,H*C,H,*NH,(diss.) + 20.51 Cal. which gives SO,'H=C,H,*NH,(diss.) + HNO,(diss.) = so c6H,< I '(dim.) + 2-0 4- 20.31 Gal. N2 The agreement of these results with those previously obtained (A. 1910 ii 588) shows that the process proceeds quantitatively independently of whether the nitrite or the sulphanilic acid is in excess and thatl tho process of diazotisation proceeds in one and the same direction in both cases.80 (2) C,H,< 1 '"(diss.) + NaOH(diss.) = SOsNa*C,H,*N,*OH(diss.) + 5.65 Cal. *2ii. 108 ABSTRACTS O F CHEMICAL PAPERS. and C6H4<~03(diw.) + 2NaOH(diss.) = hence SO3Na-C,H4*N,*OH(diss.) + NaOH(diss.) = If a large excess of sodium hydroxide is used the heat of formation of the pseudo-salt S03Na*C6H4*N,*ONa is 6-18 Cal. the mean between this value and 4-50 Cal. approximating closely to the value 5-46 Cal. previously given (Zoc. cit.). Simple Form of Micro-balance for Determining the Den- sities of Small Quantities of Gases. F. W. ASTON (Yroc. Roy. SOC. 1914 [ A ] 89 439-446).-A simple form of quartz microlbalance (compare Steele and Grant A. 1909 ii 876) is described in which a small quartz bulb of about 0.3 C.C.capacity fused to the end of one arm is balanced by a fixed counterpoise consisting of a piece of quartz rod attached to the end of the other arm. With this arrangement the pressure necessary to bring the beam to zero affords a meaaure of the density of the gas in the balance case. This consists of a cell. about 3 mm. wide made of thick glass plates accommodation for the bulb being provided by an attached glass tube which is closed by a glass plug pushed in as far as possible without actually touching the bulb itself. The case is connected by a short capillary tube to the gas-admission apparatus and pump through one stop-cock ard to tho manometer through another. The total volume of the balance case manometer and tubes through which the gm is admitted is only a few c.c.so that the quantity of gas necessary for measurements a t a pressure of about 100 mm. of mercury is only about 0.5 C.C. In making measurements with the micro-balance the gas of which the density is required is admitted into the balance case and manometer and the pressure is raised until the bulb rises and the knob at the extremity of the counterpoise appears on the field of a fixed reading microscope. The pressure is then carefully adjusted until the knob reaches some definite arbitrary zero when the pressure is read off. The gas is now pumped out and the operation repeated with a gas of known density. The ratio of the densities is then given by the inverse ratio of the pressures. The determination can be performed in a few minutes with an accuracy of 0.1%.It is suggested that the instrument may conveniently be applied to the measurement of pressure in gases of known density. N S03Na*C,H4-N2*ONa(diss.) + H,O(liq.) + 10.15 Cal. ; S03Na*C,H4*N2*ONa(diss.) + H20(liq.) + 4.50 Cal. T. H. P. H. M. D. The Relationships between the Go-volume b and the Critical Constants. L. GAY (Compt. rend. 1914 158 34-36).- The author has determined the values of the ratios B,/b and RB/nb (compare A. 1913 ii 1027) for a number of other substances in- organic and organic gaseous and liquid. For the ratio RB/.lrb the inorganic substances give values con-GENERAL AND PHYSICAL CHEMISTRY. ii. 109 cordant with that already found namely 13.84 (Zoc. cit.). The alcohols aniline benzonitrile and water however give results dif- fering somewhat widely and this the author conslders is due to their polymerisation.The ratio V,/b for which the value 3-65 had previously been obtained needs t o be increased to 3.95 for the difficultly liquefiable gases such as nitrogen and argon. The other substances give values closely approximating to 3-65. W. G. Pieaometric Analysis. JEAN TIMMERMANS (Bull. Acctd. Roy. Belg. 1913 810-83O).-A method of studying the changes which occur in condensed systems is described. This depends on the observation of the changes in volume which take place when the compressed system is allowed to return very slowly t o the condition which it assumes under ordinary atmospheric pressure the temperature being maintained constant during the process. If the pressure readings are plotted against time piezometric curves are obtained which are in many ways analogous to the well-known cooling curves of thermal analysis.The nature of the piezometric curves which are to be expected for the various types of changes which occur in one or two com- ponent systems is examined and discussed in detail. H. M. D. Viscosity of Binary Mixtures. JAMES KENDALL (Medd. K. Vet. Nobelinst. 1913 2 No. 25 1-16).-1t is shown that tlhe viscosity of binary mixtures of non-associating chemically indif- f erent liquids can be satisfactorily represented by the formula 11 =qlnl . q2n2 o r logy=n,logq + n210gq2 in which 7 is the viscosity of the mixture n1 and n2 the molar fractions of the two components in the mixture and ql and q2 the viscosities of these components. The formula has been applied to the observations of Thorpe and Rodger Linebarger Dunstan and Getman and it is found that the average percentage deviation between the calculated and ob- served viscosity is smaller than that afforded by any of the formulz which have been previously put forward.I n the case of dilute solutions the viscosity q may be expressed in terms of that of the solvent qo by the formula y = A s . T ~ in which A is a constant and x the molar concentration of the solute with reference to a fixed weight of solvent. This formula may be derived from the previous formula by assuming that one of the components of the mixture is present in large excess. It is shown that the viscosities of aqueous solutions of sucrose and carbamide and of benzene solutions of octyl hydrogen phthalates can be satis- factorily represented by the second formula over very considerable ranges of concentration.It should be noted that the above logarithmic forrnuh only differ from those put forward by Arrhenius (A. 1888 236) in the manner in which the composition of the binary mixture is expressed. H. M. D.ii. 110 ABSTRACTS OF CHEMICAL PAPERS. Viecosity of Calcium Ghloride Solutions. FREDERICK SIM EON (Phil. May. 1914 [vi] 27 95-100).-The viscosity of a series of concentrated aqueous solutions of calcium chloride has been meas- ured a t 1 5 O and 20°. The liquid was made t o pass a t constant pressure through a horizontal tube about 30 cm. long and 0.4 mm. in diameter. Curves are drawn connecting the viscosity with the concentration and it is found that these curves are not in agree- ment with Tucker’s recent measurements (A.1913 ii 3’78) of the fluidity a t 16’77O. The curve representing Tucker’s data intersects the author’s curve for 20° a t a high concentration. H. M. D. Is Poiseuille’s Law Valid for Suspensions. M. KOTHMANN (PJluyer’rr Archie 1914 155 318-348).-X form oE viscometer is described by means of which it is possible to make continuous measurements of the rate of flow of a liquid through a capillary tube when the driving pressure is increased a t a steady rate. One end of the horizontal capillary is attached to a wider glass tube serving as reservoir whilst the other end is connected t o a second glass tube of uniform diameter and also perfectly cylindrical.The position of the liquid meniscus in this tube affords a measure of the volume of the liquid which has passed through the capillary. This position is recorded on a photographic- film which moves with cbnstant speed in a direction perpendicular to the capillary tube. In the case of a liquid which obeys Poiseuille’s law the record obtained on the film for the movement of the liquid in the wider horizontal tube will be represented by a parabolic curve. This result was actually obtained in experiments with a mixture of glycerol and water. A parabolic curve was also obtained with frog’s blood and with dog’s blood (diluted wit.h serum in the ratio 1 10) when a capillary tube of a diameter equal to 465p wits em- ployed. I n both these cases quite different curves were obtained however when a tube of much smaller diameter (loop,) was used.The evaluation of these curves by a graphic method indicates that the viscosity diminishes as the pressure increases. The rate of diminution of the viscosity with increasing pressure is much greater in the case of the frog’s blood which fact is doubtless connected with the much larger size of the corpuscles. In the matter of the number of the corpuscles the two liquids were practically identical the dilution of the dog’s blood with serum in the above-mentioned ratio being such as to make the liquids strictly comparable in this respect. The experiments indicate that Poiseuille’s law is not valid for liquid suspensions moving through narrow capillary tubes. For such liquids the viscosity depends on the driving pressure and also on the ratio of the diameter of the suspended particles to the diameter of the capillary. From an examination of the available observations relating to the circulation of blood in the living organism the author arrives a t the conclusion that this does not occur in accordance with the Poiseuille’s law.H. M. D.GENERAL AND PHYSICAT CHEMISTRY. ii. 111 The Laws of Absorption of Garbon Monoxide by Blood in vitro. MAURICE NTCLOUX ( C m p t . rend. 1913 157 1425-1428. Compare A. 1913 ii 1071; and Douglas Haldane and Haldane A. 1912 i 591).-The percentage of hzemoglobin combined with carbon monoxide in blood placed in contact with a mixture of carbon monoxide and oxygen can be determined by the equation y = lOOz/z(l- K ) + IOOK where y is the percentage of haemoglobin combined with the carbon monoxide and x the percentage of carbon monoxide in the gaseous mixture referred to their respective pres- sures in the mixture and K is a constant.The author has verified this law in the case of pig’s blood and has obtained very concordant values for K on varying the values of z. Haemoglobin in blood therefore when in contact with mixtures of carbon monoxide and oxygen combines with the two gases in proportions defined by their respective pressures in the mixture and regulated by the law of mass action. The same holds good if air is substituted for oxygen it being necessary then to consider only the oxygen. in the air. W. G. Adsorption from Solutiona. A. M. WILLIAMB (Medd. K Vet. Nobelinstitut 1913 2 No.27 1-23).-Experiments have been made on the adsorption of acetic acid by blood charcoal from aqueous solutions and on the adsorption of acetic acid and water from the vapour in contact with acetic acid solutions of varying concen- tra tion. It is pointed out that both solute and solvent may be adsorbed and when this occurs the methods usually adopted f o r the measure- ment of the adsorption lead to erroneous conclusions for neither the volume of the solution nor its mass can be regarded as constant. It is therefore suggested that the concentration of the solution should be referred t o a constant weight of solvent. In cases where the adsorption equilibrium is the same indepen- dently of whether adsorption takes place from solution or from the vapour phase in contact with it it is shown that the amounts of solvent and solute adsorbed may be calculated.In terms of this double adsorption it is possible t o give a simple explanation of the phenomena of ‘‘ negative ” and Observations a t great dilutions seem to show that the value of n in the exponential adsorption equation a=kcn becomes a whole number or a simple fraction. anomalous ” adsorption. H. M. D. Utilisation of Diffusion Processes in the Preparation of Pure Substances. JOHN JOHNSTON (J. AM. Chem. Soc. 1914 36 16-19).-The formation of large particles of slightly soluble sub- stances can be secured by causing the separation of the substance to proceed very slowly and by so selecting the temperature and composition of the medium as to have the substance as soluble as possible.The simplest way of carrying out such a process is to fill with a suitable liquid a large vessel a t the opposite sides of which are placed two small evaporating dishes containing the solid sub- stances. The vessel is protected from violent fluctuations of tem-ii. 112 ABSTRACTS OF CHEMICAL PAPERS. perature in order t o prevent the disturbing effect of convection. By this means large hexagonal prisms of calcium hydroxide can be obtained from calcium chloride and sodium hydroxide a t the ordinary temperature. Large crystals of barium sulphate have been obtained in a similar manner a t looo which were practically free from the impurity occluded by this salt when precipitated in the usual way. Another method is described in which solutions of the interacting substances are placed in flasks provided with syphon tubes which dip into the main mass of liquid.Each solution is transferred very slowly since air is only admitted into each flask through a second tube which is drawn ou,t to a very fine capillary. The crystals obtained in this way are of great purity and are theref ore very suitable for the determination of physico-chemical constants. E. G. Molecular Gomplexes in Solutions and the Distribution of Certain Substances bet ween Benzene and Water. F. BUBANOVI~' (Medd. K. Vet. Nobelinstitut 1913 2 No. 33 l-BS).-Ry means of measurements of the changes which occur in the freezing points of solutions of ethyl alcohol acetic acid and chloral hydrate in benz- ene when successive small quantities of water are added to the solutions it has been found that each of these substances forms hydrates by combination with the added water.From these data and the molecular association of the three substances in anhydrous benzene solution i t has been found possible to deduce approxi- mately the composition of the hydrates. The calculations indicate that ethyl alcohol in the more dilute solutions forms the hydrate GEtOH,H,O whilst in solutions which contain a larger proportion of alcohol the hydrate 8EtOH,H,O appears to be formed. I n the case of acetic acid and chloral hydrate the complex molecules formed correspond with the formulze 4CH3*C0,H,H,0 and 4CC13*CH( OH),,H,O. The solubility of water in benzene solutions of the above three substances has also been determined.The results show that the increase in the solubility is proportional to the square of the quan- tity of ethyl alcohol associated with a definite quantity of benzene whilst the increase of the solubility in the case of acetic acid and chloral hydrate solutions is proportional t o the first power of the quantities of these substances. Other experiments have been made on the ratio of distribution of alcohol acetic acid and chloral hydrate between benzene and water. These show that all three substances form complex mole- cules in the benzene solution. H. M. D. Solubility of Crystals on Different Faces. J. LEBRUN (Bull. Acad. Roy. BeZq. 1913. 953-979).-A solution of sodium nitrate (containing 48.55 grams of _salt per 100 c.c.) dissolved from the faces ( l o l l ) (OOOl) and (1210) of a crystal of sodium nitrate 11.8 10.8 and 10.7 mg.per sq. cm. per minute respectively. A solution of magnesium sulphate (33.10 grams MgSO per 100 c.c.)GENERAL AND PHYSICAL CHEMISTRY. ii. 113 dissolved from the faces (OOl) (OlO) (110) and (111) of a crystal of magnesium sulphate (MgS0,,7H20) 3.7 3.5 3.5 and 3.7 mg. per sq. cm. per minute respectively. The amount of material in mg. per sq. cm. per day dissolved from different faces of a quartz crystal by hydrofluoric acid of concentration I (9.718 grams) and I1 (20.169 grams HF per 100 C.C. of solution) is as follows I. 11. Base (0001) ....................... 7 -2 18-76 Prism of second order (1 120) ... 1.15 4-37 Prisiri of first order (iOT0) ...... 1-17 4'41 Direct rhombohedron (1011) ...0.97 3 *7 Fused quartz ........................ - 70.0 L. J. S. Solubility Differences on Grystal Surfaces. ALBERT RITZEL (Zeitsch. ph?ysikaZ. Chem. 1913 86 106-108).-Polemical against Kuessner (A. 1913 ii 932; compare also Ritzel A 1911 ii 488). J. F. S. Modifications of the Form of Crystals of Some Substanoes Artificially Goloured during their Growth. PAUL GAUBERT (Compt. rmad. 1913 157 1531-1533. Compare A 1906 ii 152 343; 1910 ii 4 ; Marc A. 1911 ii 193 258).-Colouring matters dissolved in aqueous solutions of various salts from which crys- tallisation is proceeding enter regularly into the crystals during the growth not only modifying their faces bue also causing the formation of macles. W. G. Retger's Rule as a Criterion of Isomorphism. W. STORTEN- BEKER (Rec.t m w . chim. 1913 32 210-225).-A theoretical paper in which the various means of testing isomorphism are discussed. It is shown that there is nothing or a t most very little in modern work which is incompatible with the view that the properties of isomorphous mixtures are a continuoue linear function of the com- position; this is most markedly seen in the case of the specific volume of isomorphous mixtures. The Action of Hydroxyl Ions on Clays and Kaolins. 11. PAUL ROELANU (Bioci~em. Zeitsch. 1913 58 202-204. Compare A. 1912 ii 1150; 1913 ii 412).-Very small hydroxyl ion concentra- tions have no influence on the sedimentation of kaolin emulsions. I f the concentration is increased the sedimentation effect is also increased up to a certain maximum point.Increase of the hydroxyl concentration above this point generally slowly diminishes the sedimentation. S. B. S. J. F. S. Equilibrium in Ternary Systems. X. F. A. H. SCHREINE- MAKERS (Proc. K . Akad. Wetensch. Amstwdurtt 1913 16 540-552. Compare A. 1913 ii 851; this vol. ii 41).-A theoretical paper in which the author discusses the nature of the equilibria in ternary systems characterised by constancy of both temperature and pres- sure. H. M. D. \TOT,. CVI ii 8ii. 114 ABSTRACTS OF CHEMICAL PAPERS. Equilibria in the System Sulphuric Acid Ammonium Sulphate and Lithium Sulphate at 30". G. C. A. VAN DORP (Zeitrrch. physikccl. Chem. 1913 86 109-112. Compare Dorp A. 1910 ii 698; and Schreinemakers and Cocheret A. 1906 ii 424; 1907 ii 23).-The equilibrium relations of the four-phase system water sulphuric acid ammonium sulphate and lithium sulphate have been determined and the results plotted on a projected spatial co-ordinate system.The following hitherto unknown compounds were isolated (NH,)2S04,3Li,S04,4H2804 ; 2(NH,)2S04 2Li2SO4,4H,SO ; and 3(NH,)2S04,Li2S04,4H,S0 J. F. S. Neutral Salt Action. BOHDAN DE SZYSZKOWSKI (Medd. K. Vet. Nobelinstitut 1913 2 No. 41 1-55. Compare A 1907 ii 238; 1908 ii 761; 1910 ii 703).-The experiments described were carried out in order to determine the influence of strong electro- lyt,es on the affinity constants of weak acids. It is shown that this influence may be calculated from observations on the distribution of the acids between benzene and water containing varying amounts of a strong electrolyte.The acids used were benzoic and salicylic with sodium chloride as the electrolyte and the concentration of this was varied from 0.1 to 1N. If the values of the affinity constants of the acid are plotted as a function of the sodium chloride concentration the curve for benz- oic acid is found to rise from 6.0 x in pure water t o a maximum value of 8 . 2 ~ 10-5 a t about 0*5N. As the concentration of the sodium chloride is further increased the affinity constant falls and in 1N-solution its value is only 4.1 x 10-5. I n the case of salicylic acid a similar type of curve is obtained the affinity constant increasing from 1 . 0 2 ~ 10-3 in water t o a maximum value of about 1-65 x in 0.25N-solution then falling to 1-00 x 10-3 in 1N-sodium chloride solution.The results obtained for a pp-ticular concentration of the strong electrolyte are independent of the concentration of the acid and this fact affords strong support to the interpretation which is placed on the experimental data. I n order to account for the maximum in the value of the affinity constant a t a definite salt concentration it is supposed that the ionisation of the acid is subjected to the influence of two opposite factors. On the one hand the ions tend to increase the ionisation of the weak acid whilst the undissociated salt molecules tend tlo reduce the ionisation. The maximum affinity constant can obviously be explained on the basis of this hypothesis for the action of the ions will preponderate in the more dilute salt solutions and that of the undissociated molecules will be the more important in more concentrated solutions. Measurements of the solubility of benzoic and salicylic acids in water containing variable amounts of sodium chloride have given results which support the above view.I n this connexion it may be noted thatl the solubility of salicylic acid in 0-1N-sodium chloride solution is somewhat greater than it is in pure water. The later portions of the paper are concerned with the explana-GENERAL AND PHYSICAL CHEMISTRY. ii. 115 tion of the deviations from the mass law which are exhibited by strong electrolytes and the view is expressed that such deviations are merely due to effects comparable with that which has been examined by the author. When it is possible to apply the necessary correction for such neutral salt effects i t will probably be found that the mass law is applicable to the ionisation of all electrolytes.H. M. D. The Combustion of Gaseous Mixtures and the Velocities of Reaction. TAFFANEL (Compt. rend. 1914 158 42-45).- A mathematical discussion of results already published (compare A. 1913 ii 1039). W. G. Influence of Hydrogen Ion and of Neutral Salts on Colour Changes and Reaction Velocities among Dyes of the Triphenylmethane Series. H. C. BIDDLE (J. Anror. Chetrh. Soc. 1914 36 84-104).--It has been shown in an earlier paper (A. 1913 ii 312) that the action of an excess of an acid on a rosaniline dye or its carbinol base leads to an equilibrium between the coloured and the colourless forms which is largely dependent on the concen- tration of the hydrogen ions of the acid.A further study of this change has proved that the velocity of the reaction leading to this equilibrium is a function of the diminishing concentration of the H' ion. The rate of the reaction differs from that of the co_nversion of the cinchona alkaloids into their toxins (Biddle and Rosenstein A. 1913,- ii 492) in being independent of the nature of the acid and being consequently a function of the molecular concentrabion of the acid. The velocity increases as the temperature rires the rate of increase becoming greater with diminishing concentration of the H' ion. Neutral salts increase or decrease the rate of the reaction in presence of a weak acid in accordance with their effect in diminishing or increasing the degree of dissociation of such an acid.The neutral salt effect gives rise to a change of colour and a diminution of the velocity of the reaction corresponding precisely with the results produced by an increasing concentration of the H' ion. It has been found that within certain limited ranges the rate of reaction may be used as a dynamic method for determining the concentration of the H' ion. E. G. Expression of the Velocities of Transformation of Physico- chemical Systems as a Function of the Affinity. R. MARCELIN (Ccin7pt. rend. 1913 157 1419-1422. Compare A 1911 ii 27).- A mathematical discussion of such expressions in the case of dif- fusion,. evaporation sublimation and solution based on the hypo- thesis that a physico-chemical complex in process of transformation is constituted by two systems which evolve in opposite senses; the progressive system increases in mass whilst the regressive system diminishes.The observed velocity u is the resultant of the velocities of these systems. W. G . 8-2ii. 116 ABSTRACTS OF CHEMICAL PAPERS. Reactions of Vibriolysine. SVANTE ARRHENIUS [with THORVALD MADSEN and YUTAKA ‘rEHUUCEI] (Meclcl. h’. Vet. Nobelinstitut 1913 2 No. 39 1-39).-Experiments have been made to determine the nature of the processes which are involved in the absorption of vibriolysin by blood corpuscles coagulated serum-albumin and animal charcoal and also in its neutralisation by anti-vibriolysin. I n the case of the blood corpuscles the rate of absorption can be represented by the formula for a unimolecular reaction an equi- librium condition being reached in which the concentration of the vibriolysin in the corpuscles is 286 times that in the surrounding solution.The ratio of distribution in the equilibrium condition has been found to remain constant when the concentration is varied in the ratio 1 25. The velocity of absorption increases very rapidly with the temperature the velocity becoming 4.7 times as great for a rise of loo. The dynamic experiments with charcoal show that vibriolysin is decomposed by this substance and the data would appear to in- dicate that each particle of carbon is only capable of decomposing a definite quantity of the lysin per unit of time. I n many respects the change offers a close resemblance to reactions in which enzymes are involved.The influence of temperature on the change is relatively small a rise of temperature of loo increasing the velocity only in the ratio 1 1.5. The neutralisation of vibriolysin by anti-vibriolysin takes place in accordance with the formula for a bimolecular reaction if the concentrations of the reacting substances are reckoned from the equilibrium condition. This behaviour is closely similar to that which has been observed in the neutralisation of tetanolysin. With regard to the absorption of the lysin in coagulated serum- albumin it is found that the process leads to a condition of equi- librium and that the rate of change is approximately in agreement with the bimolecular formula. The amount absorbed in the equi- librium condition is proportional to the square-root of the concen- tration of the surrounding solution and it seems probable therefore that the phenomenon is one of absorption. It differs therefore completely from the changes which are involved in the absorption of vibriolysin by blood corpuscles and in the neutralisation by anti- vibriolysin.Although the decomposition of the lysin is the chief characteristic in the action of animal charcoal it is possible that the decomposition is preceded by adsorption. The Influence of Foreign Substances on the Activity of Catalysts. 111. Platinum a s Hydrogen Carrier. C. PAAL and E. WINDISCH (Ber. 1913 46 4010-4016).-1n continuation of previous experiments (compare A. 1913 ii 1043) the catalytic behaviour of platinum deposited on various metal powders and insoluble’ compounds has been investigated ; the particular reaction used wae the reduction of cotton-seed oil by hydrogen.The platinised metals were prepared by shaking the metal powder with an aqueous solution of chloroplatinic acid; in the case of magnesium an aqueous-alcoholic solution was used. The products H. M. D.GENERAL AND PHYSICAL CHEMISTRY. ii. 117 were washed with water alcohol and ether excepting platinised magnesium where washing with water was omitted and then dried in a vacuum. The amount of platinum deposited was 2.5% of thO weight of the metal. The compounds were coated with 2.5% of platinum by suspending them in an alkaline solution of chloro- platinic acid and subsequent reduction with hydrazine. Only magnesium and nickel have no effect on the catalytic activity of the platinum ; aluminium cobalt and bismuth diminish the activity very considerably whilst iron copper zinc silver tin and lead do away with it completely.Magnesium oxide and carbonate have no effect whereas basic lead carbonate and basic bismuth nitrate make the platinum passive. Gatalytic Esteriflcation by the Wet Method. E. BODROUX (Compt. rend. 191 3 157 1428-1 430).-A theoretical discussiou of the nature of the action in the catalytic esterification of acids by means of inorganic acids (compare this vol. i lo). The author suggests that an intermediate additive compound is formed by the organic and inorganic acids as follows T. S. P. /OH \X R*CO,H + HX = ft*C-OH where HX represents an inorganic acid. Size and Constitution of the Atoms. ADOLF HEYDWEILLER (Ann.Physik 1913 [iv] 42 1273-1286. Compare A. 1913 ii 645 9l'9).-Tbe diameters of molecules can be calculated from the mean free path in the gaseous condition and also from the molecular refractivity. A comparison of the results obtained in the cme of the monatomic inert gases shows that satisfactory agree- ment is obtained only i n thO case of xenon. With falling atomic weight the divergence between the numbers increases and in the case of helium the ratio of the values reaches 1 - 5 l . The values obtained for the diameters of the molecules of hydrogen and chlorine from viscosity data are respectively 2.176 and 3.693 x The diameters of the hydrogen and chlorine ions as deduced from refractivity data are 1.663 x 10-8 cm. and 2.765 x 10-8 cm. respectively.The ratio of the numbers for hydrogen is 0.764 and for chlorine 0.749. On the assumption that the hydrogen (or chlorine) atoms fuse together to form a spherical molecule the ratio of the atomic to the molecular diameter would be 0.794. If on the other hand the atoms remain isolated the ratio of the atomic to the effective molecular diameter would be 0.774. Both these numbers are in fair agreement with the above ratios for hydrogen and chlorine and afford support f o r the author's view that atomic dimensions can be accurately calcu- lated from refractivity data. The constitution of the atom is discussed in reference to the author's steric theory (Zoc. c i t . ) and evidence is obtained which suggesh that the number of eiectrons in the atom is about a hundred timw as large as the steric number.According to this,. the hydrogen atom would contain about 500 electrons. H. M. D. W. G. cm.ii. 118 A 13STLtACTS OF CHEMICAL PAPERS. Theory of the Electrical and Chemical Forces in the Atom. A. BYK (Ann. Physik 1913 [iv] 42 1417-1453).-A theoretical paper in which the author discusses the constitution of the atom and the nature of the forces which come into play in atomic systems. The theory is applied in the calculation of molecular dimensions values being obtained for the diameters of the molecules of the alkali metals the halogens and compounds of carbon sulphur and chlorine which are in accordance with those derived from experimental data. The heats of dissociation of chlorine bromine iodine sulphur and phosphorus are also deduced from a formula in which electro-magnetic data only are involved the values obtained being in all cases in fair agreement with the thermal data recorded in the literature.A further conclusion to which the author’s theory leads is that the forces operating within the atom give rise to a definite atomic configuration in the case of the atoms of carbon compounds and complex inorganic compounds but that such a condition of atomic equilibrium is not characteristic of the atoms of electrolytes. H. M. D. Calculations of Frequency and Certain Relations between Elastic and Thermal Constants of Solid Elements. C. E. BLOM (Ann. Physik 1913 [iv] 42 1397-1416).-The various formule which have been put forward for calculating the frequen- cies of atomic vibration are compared by reference to the values obtained for a large number of metals.The values are not in agreement and it is not possible to decide which of the formulae gives the best results. The formula comparison suggests that. the quantities 1 / E T,/ V R,/3aV and C,,/3aB are proportional to one another E being the compressibility T the absolute melting temperature V the atomic volume R the gas constant C the atomic heat at constant volume and a the coefficient of linear expansion. If this propor- tionality is assumed a number of relations can be shown to hold between 3he elastic and thermal constaats ; these relations have been obtained previously in some cases empirically and in others by theoretical deduction. The atomic frequencies show a periodic variation with the atomic weight a relation which is not surprising in view of the periodic connexion between the elastic properties and the atomic weight and of the fact that the atomic frequencies are closely related to the cohesive forcm.Although the frequency formulze are not in general applicable to compounds certain relations which these formulae suggest are to be found between the elastic and thermal properties in certain cases. For oxides of the type RO i t is shown that there is a parallelism between the hardness and the atomic heat. H. M. D. Linking of Atoms particularly those of Carbon. A. F. HOLLEMAN (Em. f m v . chim.. 1913 32 175-183).-The author evolves a theory of the linking of atoms which combines some ofGENERAL AND PHYSICAL CHEMISTRY. ii. 119 the views of Werner with those of Stark.Making use of Stark’s idea of valency electrons it is shown how the linkings can move their relative positions on the surface of the atoms. The hype thesis is made use of to explain the Walden inversion racemisation and the Baeyer strain theory. It is shown that easy formation of double linkings cannot be directly explained on the present hypo thesis. The theory is examined in the case of other atoms; the positions of the valency electrons a t opposite ends of a diameter of the oxygen atom ar0 regarded as producing a strain in the carbonyl group which explains the ease with which the carboxyl group enters into combination. The isomerism of oximes is explained if the quinquevalent character of nitrogen is admitted. The equilibrium positions of the valency electrons for elements of valencies one to six are deduced.J. F. S. Mechanical Processes of Combination. NATHANIEL T. BACON (J. Physical Chem. 1913 17 762-768).-A theoretical paper in which an attempt is made to explain certain reactions between gases and solids by the assumption that condensed layers of gas molecules are formed a t the surface of the solid in consequence of the imperfect elasticity of the molecules. P. DE KEEN (Bull. Acad. Roy. BeZg. 1913 667-679).-The author con- siders that the electron theory does not furnish any satisfactory explanation of the behaviour of elemente of variable valency. H. M. D. The Electron Theory and the Conception of Valence. H. M. D. Automatic Mercury Pump. G. P. PAMFIL (J. Chim. P h w 1913 11 801-804).-A description of a mercury pump the auto- matic action of which is maintained by the use of a water-pump. The reservoir into which the gases from the apparatus to be exhausted expand and from which they are then expelled by the mercury is relatively large so that the rate of exhaustion is rapid.The pump contains an arrangement for collecting the exhausted gases for analysis. T. S. P. Simple Stirring Arrangement for Test-tubes. H. E. SCEAUMBURG (Chern. Zed. 1913 37 1581).-h bulb the diameter of which is about 1 mm. less than the internal diameter of a te& tube is blown on the end of a piece of glass tubing. I f this is immersed in the liquid in the test-tube and then moved rapidly up and down taking care that the bulb does not come above the surface of the liquid it acts as a very efficient stirrer. T. S. P. An Electric Vacuum Oven for High Temperatures. JOHANNES WOLF and ERICH MULLER (Zeitsch,. Elektrochem. 191 4 20 1-4).-An electric oven is described which by the use of carbon resistance tubes gives a temperature of 2850O. ThO energy con- sumption for this temperature is 15 kilowatts a quantity about 30% less than that consumed by other ovens of the same type.ii. 120 ABSTRACTS OF CHEMICAL PAPERS. The furnace is fitted with an arrangement for removing the experimental substance whilst i t is a t its highest temperature. J. F. S. A Rapid Apparatus for Ether Extractions. J. J. L. ZWJKKER (Pharm. Weakblad 1914 51 21-23)-A new simple form of ether- extractmion apparatus in which the extraction vessel is immersed in a condenser the ether vapour is carried by a bent tube to the lower end of it and the condensed ether rises through the liquid to be extracted and passes through a side-tube back into the heating flask. A. J. W. A New Form of Extraction Apparatus and Reflux Con- denser. PROBODHA CHUNDRA CHATTOPADHYAY (J. Soc. Chem Ind. 1913 32 1145-1146).-The extractor consists of an outer glass jacket and an inner glass tube (thimble) of smaller diameter. The latter is provided with a series of holes at the top and with some near the middle; there is also a very small hole a t the bottom of the thimble to allow the liquid to drain away completely when the extraction is finished. There is no syphon tube as in the Soxhlet apparatus. The substance to be extracted is placed directly into the thimble or else in a fat-free paper thimble which is then placed in the glass thimble. During the extraction the liquid overflows continuously through the holes near the middle of the thimble. I n the reflux condenser the inner tube consists of a series of four bulbs with flat spirals interposed between the second and third and between the third and fourth bulbs. The advantage of this arrangement is that a central draught which is a drawback even with double surf ace condensers is prevented. A. W. DAVISON (J. Phyrrical Chem. 1913 17 737-754).-The paper contains an account of a series of experiments on the dyeing of wool and cotton which were undertaken with the special object of collecting data which might be utilised in devising lecture experiments to illustrate the theory of dyeing. The experiments were made with basic and acid dyes and the behaviour of these in alkaline neutral and acid solution was compared. Other observations have reference to the influence of the addition of sodium sulphate and the behaviour of lakes. T. S. P. Experiments in Dyeing. H. M. D.