61 General and Physical Chemistry. Spectroscopic Observations at Very High Temperatures. RAFFAELO NASINI and FRANCESCO ANDERLINI (Gazzetta 1906 36 ii 561-570).-With the help of Helbig’s carbon tube heated in an electric furnace the authors have examined the spectra of iodine and nitrogen at high temperatures. The former shows bands of a luminous spectrum distinctly at a temperature somewhat higher than 1000° and with nitrogen a luminous line and band spectrum is obtained at about 3000O. T. H. P. The Ultra-violet Phosphorescence Spectrum of Fluorspar. Variations in the Phosphorescence Spectrum of the same Element in the same Diluent. GEORGES URBAIN and C. SCAL (Compt. rend. 1907 144 30-32. Compare this vol. ii 3).-In fifteen specimens of fluorspar the presence of rare earths has been recognised by the appearance in their cathodic phosphorescence spectra of ultra-violet bands characteristic of gadolinium.On comparison of the gadolinium spectra of natural and synthetical fluorspar striking differences were observed in the intensity of bands of the same wave- length. These are due to the greater diminution in intensity of one group of bands as compared with that of the other group when the proportion of gadolinium is decreased. This was shown by preparing the fluoride from pure chalk and less than one-thousandth part of gadolinite; its spectrum is identical with that of natural fluorspar. This phenomenon is observed with other phosphorescent systems (compare Abstr. 1906 ii 138) and is termed the phenonaenon of dilution. E. H. Photochemistry and the Phase Rule.WILDER D. BANCROFT (J. Physical Chem. 1906 10 721-728).-The displacement of the equilibrium. in systems under the influence of light is considered from the standpoint of the phase rule. The variability of the intensity of the light introduces another degree of freedom. Strictly speaking there are as many degrees of freedom as there are kinds of active light but for most purposes a beam of light may be treated as though it were homogeneous. If the independent variables of a system are the n components the pressure the temperature and one kind of active light the coexistence of n + 3 phases will constitute an invariant system. Reference is made to the equilibrium between oxygen and ozone. For a given initial quantity of oxygen a t agiven pressure and temperature the equilibrium percentage of ozone is fixed if these are the only variables; it is not fixed however if ultra-violet light ia introduced as an independent variable.The case of the displacement of the equilibrium between two modifications of a light-sensitive substance in solution is also discussed. H. M. D. VOL. XCII ii. 562 ABSTRACTS OF CHEMICAL PAPERS. Radium. ALBERTO MAYORAL OLIVER (Anal. li”is. Quim. 1906 4 132-1 35).-Actinographs obtained by means of the rays from an incandescent lamp of 5-candle power were found to have blurred outlines whilst those produced by the agency of 5 mg. of radium bromide had sharply-defined contours. The blurring is regarded as an effect due to the secondary rays. BERTRAM B. BOLTWOOD (Amer. J.Sci. 1906 [ivj 22 537-538).-The fact that the experi- mentally determined rate of production of radium from uranium is very much smaller than that indicated by theory is supposed to be duo to the existence of a relatively slow-changing intermediate product and the author suggests that this is actinium. Measurement of the rate of production of radium emanation in the actinium separated from a kilogram of carnotite has furnished numbers from which the time required for the decay of the activity of radium to one-half its initial value is about 3300 years This is of the same order of magnitude as the most recent estimate made by Rutherford and this result is supposed to afford evidence in favour of the view that actinium is the intermediate disintegration product between uranium and radium.G . T. M. Production of Radium by Actinium. H. M. D. Radioactinium. OSKAR HAHN (Phil. Mctg. 1907 [ vi] 13 165-180. Compare Abstr. 1906 ii 323).-Radioactinium is a direct disintegration product of actinium and is the direct parent of actinium X . The transformation of actinium into radioactinium is rayless. The latter product emits only a-rays and is half transformed in about 19.5 days. The methods of separation employed for the isolation of radioactinium throw no definite light on its properties. J. C. P. The Supposed Derivation of Radium from Uranium. JOSE MuRoz DEL CASTILLO (Anal. Pis Quirn. 1906 3 60-64).-The experimental results obtained in the examination of the radioactive pyromorphite of Issy l’Ev&que the radioactive barytes of Carlsbad and the radioactive deposits of the thermal springs of Battaglia Baden- Baden and Nauheim (all of which materials are quite free from uranium) favour the view that radium thorium and uranium are three independent radioactive chemical elements.G. T. M. Relative Activity of Radium and Thorium Measured by the 7-Radiation. A. 5. EVE (Amer. J. Sci. 1906 [iv] 22 477-480. Compare Abstr. 1906 ii 593).-1t is found that radium is 6.9 x lo6 times as active as thorium when both are in radioactive equilibrium when the activity is measured by the y-rays. I n the preparation of the commercial salts of thorium about one-half of the radiothorium in the radioactive minerals is abstracted. By measurement of the y-activity it is found that the thorium and its products in thorianite is about 2.5 times as active as the thorium and its products in thorium nitrate.This result is in good agreement with those obtained by the a-ray method and by the emanation method.GENERAL AND PHYSICAL CHEMISTRY. 63 In consequence of the continued increase in the proportion of radiothorium commercial thorium nitrate cannot be used as a standard for 7-ray 4easureruent a8 was suggestod previously. H. 31. D. Velocity and Energy of the a-Particles from Radioactive Substances ERNEST RUTIIERFORD (PhiZ Mug. 1907 [ vi] 13 110-117. Compare Abstr. 1906 ii 719; Rutherford and Hahn ibid ).-The initial velocities of expulsion of the a-particles from the products of the radio-elements all lie between 1.56 x lo9 and 2.25 x 109 cm. per second. The a-particles emitted from thorium C have the greatest velocity and those from uranium and radium have the least velocity.The average velocity and average energy of the a-particles from the thorium and actinium families are nearly eqnal and are greater than the corresponding values for the radium family. The total energy liberated in consequence of the successive disin tegrations of a radium atom is less than the corresponding value for the thorium atom but is greater than that for actinium. It is noteworthy that for n large number of the radioactive products the velocity of expulsion of the a-particle for each of the radioactive families increases progressively as the period of transformation decreases. Hence it is probable that tho velocity of expulsion of the particle is least for the most stable atoms J.C. P The Radioactivity of Spanish Medicinal Springs JosB Mufioz DEL CASTILLO (Anal. Pis. Quim. 1006,4 119-120,147-149). -The waters from four teen medicinal springs obtained from nine hydropathic establishments have been examined with the result that several samples formerly supposed to possess no radioactive properties have now been found to exhibit a slight degree of radioactivity. Another series of investigations carried out with Engler and Sieve- king’s apparatus showed that specimens from fourteen different springs were all with one exception more or less radioactive. As these samples lose their radioactivity very appreciably on keeping the tests were made as soon as possible after collection. I n the case of the water from Burgas de Orense the dissolved gases were also very radio- active. G.T. M. The Radioactivity of Medicinal Springs containing Nitrogen. J o d M U ~ O Z DEL CASTILLO (Anal. Fie. Quinz. 1906 4 189-192).- Those Spanish mineral waters containing nitrogen which are reputed to have a valuable therapeutic action owe this property to the presence of a radioactive emanation. Samples from ten different sources all exhibited radioactivity; these specimens contained some carbon dioxide but oxygen was mor0 or less displaced by nitrogen. G. T. AT Probable Relationship Subsisting between the Radio- activity of Minerals and Cold Springgl in the Sierra de Guadarrama. JOSE Musoz DEL CA~TILLO (And. Pis. Quim. 1906 4 263-269).-A study of the radioactive minerals and springs found in the vicinity of Guadarrama leads to the view that the water owes its 5-264 ABSTRACTS OF CHEMICAL PAPERS activity to the fact that it has percolated throughlayers of radioactive minerals.These minerals belong to many different species all of which are rare ; titanium and iron are the predominant elements accompanied in certain instances by manganese bismuth copper and other common elements. The most radioactive specimens show traces of uranium. These very active substances were found on the highest points of the Sierra so that the melted snow at once comes into contact with them. G. T. M. Radioactivity of Thermal Mud deposited from the Bagni di Lucca (Tuscany). GIUSEPPE MAGRI ( A t t i R. Accad. Lincei 1906 [v] 15 ii 699-704j.-The author has separated the constituents of the mud deposited by the thermal springs of Bagni di Lucca by means of the ordinary group precipitants and has tested the radioactivities of equal weights of the various precipitates by means of the electro- scopic method. In this way he has demonstrated the presence of several radioactive substances and the physical examination of the emanations are in accord with the chemical properties of the groups in indicating the presence of radium and thorium. The sulphide pre- cipitate of Group I1 exhibits strong dispersion which is possibly due t o polonium or one of the radium products.No conclusion can be drawn as regards the presence of actinium. T. H. P. The Radioactivity of Ashes from the Last Eruption of Vesuvius [April 19061. JosE Mu3oz DEL CASTILLO ( A n d Pis.Quirn. 1906 4 l24),.-A specimen of Vesuvian ash collected from the deck of a steamer stationed a t Naples a t the time of the eruption was found to be a light grey powder which when sifted had D 1.20 and when examined by Elster and Goitel’s method indicated less than 18-57 volts per 100 grams of material per hour G. T. M. Radioactive Minerals from Motril (Granada). JosE Mu3oz DEL CASTILLO (And. Pis. Quim. 1906 4 46-48).-Samples of minerslo containing iron and copper from Motril (Granada) were generally found to be more or less radioactive as was also in a slight degree a sample of spring water from the same locality. G. T. M. Radioactive Cinnabar from Granada [Spain]. Josh MuRoz DEL CASTILLO ( A n d . Pis. Quim. 1906,4 58-59).-1n examining the radioactivity of the non-uraniferous Spanish minerals a small specimen of cinnabar from Albufiiol Granada was studied and found t o produce an appreciablo effect on a pliot,ographic plate after twenty to thirty- two days.G. T. M. Radioactivity from the Human Body. Josh Mv5oz DEL CASTILLO ( A n d Pis. Quim lt)06,4,202-205).-The rate of decay of the radioactivity exhibited by the medicinal water of Caldas de Oviedo was determined at frequent intervals (four to seven days) and i t was found that it had practically disappeared iii one-and-a-half months. Fresh samples of this water were administered to a human subjectGENERAL AND PHYSICAL CHEMISTRY. 65 with the result that a slight amount of radioactivity could be detected by Engler and Sieveking's apparatus in the urine and also in the expired air.G. T M Electrical Conductivity of Alloys. I. Relation between Conductivity and Constitution. W. GUERTLER (Zeit. unorg. Chem. 1906 51 397-433).-From a comparison of the curves representing the relation between the electrical conductivity and composition of numerous binary alloys with the constitution of tlhe latter as deter- mined more particularly by thermal and microscopic methods a number of rules showing how the constitution of alloys can be deduced from their electrical conductivity have been established. The available data on conductivity and on constitution are discussed in detail and it is shown that the rules in question give results which are trustworthy and in good agreement with those obtained by other methods when account is taken of the very different temperatures at which the conductivity and thermal observations have been carried out.The conductivity data are taken from papers by Matthiessen Le Chatelier and others and are compared with the results of thermal and microscopic observations by Tammann and his pupils by Heycock and Neville and others. The more important rules connecting conductivity and constitution of binary alloys are as follows (1) Alloys the electrical conductivity of which is a linear function of the volume concentration of the components are not mutually miscible to any appreciable extent in the solid state and conversely ; (2) the conductivity curve (the values of the conductivity being plotted as ordinates against the volume con- centration) of alloys which form a complete series of mixed crystals is continuous but falls rapidly on both sides from the points representing the conductivity of the pure metals and shows a flat minimum; (3) if the components show only limited miscibility the first law holds for the parts of the curve between the concentrations of the saturated mixed crystals the second law for the parts between the pure metals and the saturated mixed crystals ; (4) if two metals form m compounds the complete diagram can be divided up into m + 1 single binary diagrams and conversely from the shape of the curve information as to the existence of compounds can be obtained It is pointed out that a sharp point on the conductivity curve always indicates the presence of a chemical compound but contrary to the view of Liebenoff the converse does not always hold.I n the course of the paper the effect of small quantities of other metals on the properties of iron are discussed and it is shown in accordance with the above laws that the conductivity of the iron is only seriously affected when the admixture is present in solution. From an investigation of the system copper-zinc Shepherd (Abstr. 1904 ii 662) had drawn the conclusion that the elements do not enter into chemical combination whilst the conductivity curve as determined by Matthiessen indicates formation of a compound The author points out that Shepherd's observations may be interpreted as indicating the existence of a compound of the formula Cu,Zn3. A bibliography of the subject is appended. G. S.68 ABSTRACTS OF CHEMICAL PAPERS.Variation of Electrical Condtlctittity with Change of Temperaturd id Solutions of Sodium and Potassium Chlorides. AURELIO SUAREZ INCL~N (Anal. Pis. Quim. 1906 4 94-98).-The electrical condiictivities of solutions of 1 mol. of sodium chloride in 17.73 26.52 and 35.36 mols. of water were determined over a range of temperatures from 2-96' ; the results when plotted Rhowed points of inflexion for the first solution a t 20° and for the other two at 35'. Solutions containing 1 mol. of potassium chloride in 11.85 23.70 35.55 and 47.40 mols. of water were similarly examined at 3-94'; the first solution seemed to have a point of inflexion below Oo whilst with the other three the inflexion occurred at 50° SO0 and 7 5 O respectively G. T. M.Relation of Solution Pressure to Surface Condition In Metals. CHARLES E. FAWSITT ( J . SOC. Chenz. Ind. 1906 25 1133-1134).-The surface condition of rods of iron silver and gold has been examined by measuring the difference of potential between these and solutions of the corresponding salts. By rotating the rods a t about six revolutions per second constant results were readily obtained. With a series of steels containing different amounts of carbon i t was found that the potential difference in 0.5 molar ferrous sulphate solution is practically constant and equal to that of ferrite. The solution pressure of brightly polished silver is diminished by annealing a t a red heat but attains its initial value after the silver has been hammered until hard. I n the case of gold after annealing and hardening the solution pressure is greater than that of the polished metal.A roughening of the surface of iron or steel by etching or blasting also produces an increase of the solution pressure. This is attributed to the greater surface of contact of the metal with the solution. H. M. D. The Hydrogen-Oxygen Cell. 11. FRITZ HABER and GLYN W. A. FOSTER (Zeit. anorg. Client. 1906 51 259-314. Compare this vol. ii 6)-In a former paper it was showh that whilst the -&'.M.F.'s of hydrogen and oxygen concentration cells and of hydrogen- oxygen cells at high temperatures gold and platinum being used as electrodes and glass and porcelain as electrolytes are in satisfactory quantitative agreement with the thermodynamical theory the influence of varying pressures of water vapour on the E.M.F.of hydrogen concentration cells is in qualitative but not in quantitative agreement with the theory. It is now shown that this divergence is to some extent at least due to the absorption and retention of water vapour by the material of the electrolytes even a t 1000'. I n the experiments with porcelain the middle part of a tube of this material was coated outside and inside with platinum or gold; the outer electrode was in contact with a slow stream of air and served as normal electrode whilst the inner electrode was in contact with the gases under investigation. The experiments at 860" and 1000° con- firmed the results of Haber and Fleischmann (Em. cit.). As regards the influence of water vapour the thermodynamic theory requires that if the material of the electrolyte is not indifferent to it,GENERAL AND PHYSICAL CHEMISTRY.67 the effect of increased vapour pressure of water on the E.M.F. of the OXygeh concentration cell must be in the opposite direction to that in the hydrogen concentration cell and for the same change of vapour pressure the magnitude of the effect should be equal in the two cases. Experiments with porcelain show that the changes of E.M.F. are ih opposite directions as is to be expected but a slight divergence is found from the values required by theory probably due to experi- mental error. Measurements with glass were made at 450" and it is shown that the deviations from the simple formula for the E.M.F. are due to the action of water vapour on the electrolyte.G. S. The Hydrogen-Oxygen Cell. 111. FRITZ HABER (Zeit. anorg. CItem. 1906 51 356-368. Compare preceding abstract).-The paper deals with the development of a suggestion which is now finding fairly general acceptance (compare Nernst and von Wartenberg Abstr. 1906 ii 729 ; Lorenz and Hauser Abstr. 1906 ii S25) that a platioum electrode saturated with oxygen is not completely reversible an oxide of platinum being formed which is electrically active but has a smaller oxygen pressure than the gas itself. From I3ose's value for tho E.M.F. of the hydrogen-oxygen cell (1.14 volts) as compared with the thermodynamical value at the ordinary temperature 1.23 volts it is calculated that the oxygen pressure of the compound in question is I n a previous paper the reduction potential of hydrogen peroxide was estimated a t 0.8 volt and from this employing the former value for the E.M.F.of the hydrogen-oxygen cell 1.4 volts was obtained as the oxidation potential of the peroxide. With the new value 1.23 volts the oxidation potential of hydrogen peroxide is 1-66 volts and is therefore identical with the potential at which water is decomposed between platinum electrodes. On this basis it is suggested that in the latter process two O H ions unite to form hydrogen peroxide which is immediately decomposed in contact with the oxidised platinum surface. If a platinum electrode is dipped into hydrochloric acid containing chlorine and then washed thoroughly with miter and hot alkali it gives a blue coloration with potassium iodide and starch much more intense than that produced by the catalytic action of the metal on the oxygetl of the air.Since this effect is retained when the potential falls to that of platinum in contact with a solution containing Pt"" ions it is considered that the metal has been attacked with formation of a solid platinum compohnd. An electrode which has undergone anodic polarisation in sulphuric acid shows similar behaviour. These o bserva- tions also support the view that passivity is connected with the forma- tion of a superficial layer of oxide. In addition to the investigations at high temperatures with glass and porcelain as electrolytes (preceding abstract) similar measurements have now beeli carried out with potassium and sodium hydroxides as electro- lytes and it is shown that in this case also the results are in excellent accord with the thermodynamical theory.atmospheres. G. 8.68 ABSTRACTS OF CHEMICAL PAPERS. Electrolysis of Dilute Solutione of Acids and AJkaIiE) at Low Potentials Dissolving of Platinum at the Anode by a Direct Gurrent. GEORGE SENTER (Trams. Faraduy Soc. 1906 2 142-149).-Dilute solutions of sulphuric acid and sodium hydroxide have been submitted to the action of a current between platinum electrodes at a potential below that at which gaseous oxygen is evolved. The current density a t the anode was 1-2 x 10-7 amperes per square centimetre and the anode potential -1-4 to -1.55 volts the electrolysis being continued for three or four days. I n these circumstances an oxidising substance is formed in very small quantity at the anode.T t is very stable especially in alkaline solutions and is not destroyed by boiling. It liberates iodine from hydriodic acid but does not react with titanic acid like hydrogen peroxide. Under the same conditioxis small quantities of platinum are dissolved from the anode in acid solu- tions this action taking place to a greater extent with electrodes which have been used for some time then with fresh electrodes. The solvent action is not observed in the electrolysis of alkalis and the author supposes that the platinum is superficially oxidised the oxide dissolving in acid but not in alkaline solutions. H. M. D. Electrolysis of the Alkali Chlorides. Conductivity Density and Speciflc Heat of Simple and Mixed Solutions of Sodium Chloride and Sodium Hydroxide.LOUIS DEMOLIS (J. Chim. Phys. 1906 4 528-546).-Numerous data have been obtained for the electrical conductivity density and specific heat of solutions containing sodium chloride and hydroxide at a series of different concentrations and temperatures. The density of pure sodium chloride solutions and of mixed solutions can be expressed by the equation Dt = [0.999871+ 0*037298N- 0.O62455N2 + 0.00103R - 0*06688122]/1 + 0.000398 in which N and R are the numbers of grams of sodium chloride and sodium hydroxide per litre respectively and t is the temperature. The specific heat values for sodium chloride solutions are found to satisfy the formula proposed by Mathias. H. AT. D. Electrolysis of the Alkali Chlorides. Refractive Index Viscosity and Ionic Transport Ratio of Simple and Mixed Solutions of Sodium Chloride and Hydroxide.E. BRINER (J. Chim. Phys. 1906 4 547-564).-From the experimental data for the refractive index at 18' and for the sodium line the following equations are deduced. For solutions of sodium chloride n = 1.3334 + O.O31683C - O-O669C2; for sodium hydroxide n = 1.3334 + 0.032694C - 0*0628YC2 ; for potassinm hydroxide n = 1.3342 + O*O31866C- O.OGl12C2 in which C denotes the concentration in grams per litre. The refractive index of a solution containing sodium chloride and hydroxide cannot be represented accurately by an equation of this The data for the coefficient of viscosity a t 20°.can be represented by similar equations. Sodium chloride q = 0.010015 + 0-O41O04C + 0.0764C2 ; sodium hydroxide q = 0.010015 + O,O4457C + O*OG3321C2 ; potassium hydroxide y = 0*010015 + O.O418C + O*O74C2.The viscosity type.GENERAL AND PHYSICAL CHEMISTRY. GD coefficients for solutions of sodium chloride and of potassium hydroxido art! also calculated by means of the formula suggested by Euler. The transport number of the anion in sodium hydroxide a t a concen- tration of 84.2 grams per litre was found to be 0-85 a t 20-25' and 0.858 at 15O; in potassium hydroxide a t a concentration of 152 grams per litre 0.75". H. 31. D. Electrolytic Oxidation in presence of Fluorine Ions MARIO G. LEVI and F. AGENO (Atti R Accad. Lincei 1906 15 ii 549-555 615-620. Compare Skirrow Abstr. 1903 ii 69 and Miiller Abstr. 1904 ii S l l and Sl2).-Using solutions of chromium sulphate normal as regards the salt and sulphuric acid the authors find that the yield of chromic acid obtained on electrolysis is increased by the presence of hydrofluoric acid and by the use of platinised platinum electrodes.The highest yield obtained was 78% of chromic acid a value comparable with that obtained industrially by means of peroxidised lead electrodes (compare Ahrens Ekktrochemie 1903 539 and Muller and Soller Abstr. 1906 ii 66) ; the amount of hydrofluoric acid present 0*498N mas only about one-fifth of that used by Skirrow (Zoc. cit.). The diminution of the current density due to the platinising of the electrodes probably everts a favourable influence on the yield of chromic acid. The addition of sodium fluoride favours the electrolytic oxidation of sodium sulphite to dithionate with which corresponds a higher anodic potential than with the sulphate the other oxidation product of the sulphite.With smooth platinum electrodes the yield of dithionate is raised from 18.6 to 21.3% by the presence in the solution of 1.25% of sodium fluoride and from 21 to 26.1% by 3% of the fluoride. When platinised platinum electrodes are used the presence of fluorine ions is insufficient to raise the anodic potential to the value necessary for dithionate to be formed so that the sulphite undergoes complete oxidation to sulphate. The authors have also electrolysed either solutions of ammonium sulphate rendered alkaline by sodium or potassium hydroxide or solutions of free ammonia in presence of sodium or potassium fluoride the electrodes being of smooth platinum and the cathode being Eeparated by a diaphragm containing 20% sodium or potassium hydroxide solution according to the fluoride used.The oxygen and nitrogen (by difference) in the anodic gases were determined and also the nitrite and nitrate formed in the solution. I n this oxidation also the fluorine ions produce a slight but appreciable increase in the yield of the compound with which corresponds the higher potential that is the nitrate. Electrolysis of manganese salts in presence of hydrofluoric acid gave results agreeing with those of Skirrom (Abstr. 1903 ii 69). For the estimation of the permanganate the authors used a method similar to Ballmann's spectroscopic method for estimating lithium in mineral waters (compare Ranzoli Abstr.1901 ii 423) the appearance or disappearance of the characteristic striae in the absorption spectrum of the permanganate being observed Using an h u e r light and a layer of solution 10 cm. in thickness the minimum concentration of70 ABSTRACTS C1F CHkMICAL PAPERS. permanganhte for which these s t r i a are visible is 0.000002 gram per C.C. The use of fluorine ions was also applied i n the electrolysis of a number of organic compounds but no useful results were obtained. T. H. P. Specific Heat and Speciflc Gravity of Allotropic Modi- fications of Solid I Elements. ALBERT WIGAND (Ann. Physik 1907 [iv] 22,64-98).-From Kicharz’s theory (Abstr. 1893 ii 404 ; Ann. Pliysik 1899 67 704) it is possible to deduce the conclusion that the greater the specific gravity of any modification of an element the smaller is its specific heat. It is now shown partly with data obtained by the author that the rule is valid for the various modifications of carbon boron silicon phosphorus sulphur arsenic selenium tellurium and tin.J. C. P. Variation with Temperature of the Specific Heat of Solid Elements. ALBERT WIGAND (Ann. Phgsik 1907 [iv] 22 99-106). -Available data bearing on the relation of the specific heats of the elements to temperature are brought together and represented on a large diagram. The extent to which elements of low atomic weight and small atomic volume deviate from Dulong and Petit’s law is discussed and it is shown that the validity of the law is much more extensive at high temperatures than at low temperatures. J.C. P. Phenomena accompanying Fusion and Crystallisation. DANIEL VORLANDER (Zeit. phpikal. Chem. 1906 57 357-366).- - Under normal conditions p-anisylideneanisidine p-anisylidenephenet- idine and anisylidene-p-aminoacetophenone melt to isotropic liquids but if drops of the liquid substances are supercooled a fluid aniso- tropic phase appears exhibiting double refraction. The supercooling of the isotropic liquid is in fact of great importance in the search for fluid crystalline substances The following substances are capable of existence in two fluid crystal- line phases (compare Lehmann Abstr. 1906 ii 836) anisylidene- aminoacetophenone ethyl p-azoxybromocinnamate ethyl anisylidcne- p-aminocinnamate and ethyl p-acetoxyazobenzeneacrylate. There are therefore for each of these substances three temperatures which correspond with a freezing point or a transition temperature thus solid crystalline bright fluid crystalline dark fluid crystalline isotropic liquid Another interesting substance is anisylidene- p-aminobenzoic acid which exists in two solid modifications and also inelts to a crystalline fluid.The author has succeeded in observing fluid crystals with straight edges and regular angles and the photographic reproductions accom- panying the paper show that there is but little difference in the mode of growth of fluid and solid crystals. It is noteworthy that the form of fluid crystals is affected to a relatively slight extent by alterations in chemical composition 1 2 3 J C. P,GENERAL AND PHYSICAL CHEMISTRY.71 Apparatus for Sublimation in a Vacuum. R. KEMPF (CAem Zeit. 1906 30 1250. Compare Krafft and Weilandt Abstr. 1896 ii 635 ; Kiiber Abstr. 1900 ii 468).-The apparatus described and figured consists of three glass parts a long bulb which is inclined downwards and in which the substance is heated is ground into the end of a wide horizontal tube which receives the sublimate; the other end of the tube is ground into a cap fitted with a tap. The nozzle of the cap is attached to a pump and the bulb and a portion of the tube are inserted into an air-oven which is heated after the apparatus has been evacuated. The advantages claimed for the apparatus are that the sublimate cannot drop back on t o the heated surface that the sublimation may be carried out under any desired pressure or in the cat.hode-light vacuum and that the sublimate can be removed quantita- tively and the apparatus readily cleaned.G. Y. Distillation and Desiccation in Vacuum by Means of Low Temperatures. A R S ~ N E D’ARSOKVAL and FRED. BORDAS (Ann. C’him. anal. 1907 12 4-7).-The distilling flask or desiccator is connected with a water air-pump the vapours are condensed on a surface cooled by liquid air or solid carbon dioxide and acetone and the last traces are absorbed by especially prepared charcoal. A Crookes’s vacuum bulb serves as manometer. For details the original article and illustration should be consulted. L. DE K. Ebullioscopic Behaviour of Aliphatic Acids with Abnormal Vapour Densities. ERNST BECKMANN [with E. BERNHARD EREMIE- POPA and WERNER GABEL] (Zeit.physikal. Chern. 1906 57,129-146). -The boiling point elevation constant Ii can be calculated by the formula R= 0*025!2/w even for those solvents (formic acetic propionic and butyric acids &c.) the vapour densities of which are abnormally large a t their boiling points. The following table gives the mean values of Ji for the four fatty acids mentioned obtained experimentally by using such solutes as bend benzanilide and diphenylamine and also the values of K calculated from the latent heat of vaporisation I{. From b. p. K. From heat Solvent. experiments. of vaporisation. Formic acid ............. 24.0 23.2 Acetic acid ............... 29.9 30.7 Propionic acid ............ 35.1 37.5 Butyric acid ............... 39.4 33-4 The available data for the heat of vaporisation in manycases exhibit marked discrepancies and in these cases at least the experimentally deduced values of K are to be preferred.Sodium and potassium formates and potassium sulphate appear to be dissociated to a large extent in formic acid solution. Acetates pro- pionatea butyrates and stearates have in general the normal moleculsr weight when dissolved i D the corresponding acid but it is noteworthy that strontium and calcium acetates exhibit a tendency to association in acetic acid solution. J. c. P.72 ABSTRACTS OF CHEMICAL PAPERS. Thermostat for Low Temperatures. W. KUNTZE (Centr. Bakt. Par. 1906 ii 17 684-688).-A thermostat for low temperatures (about 20') is described with sketches. It consists of a double-walled wooden box containing a lower zinc water chamber which may be heated when necessary and an upper one for cooling.When desired the water in the lower chamber can be quickly cooled by means of a cooling tube connected with the upper cold water chamber. Behaviour of Certain Substances at Low Temperatures. A. HEIDUSCHKA ( A ~ c h . Phnwn. 1906 244 569-571).-The following miscellaneous observations at the temperature of liquid air - 186" are selected from those enumerated. Alcoholic solutions of indigotinsulphonic acid and of certain sulphur dyes fluoresce. The fluorescence of an alcoholic solution of methyl-violet BB was not diminished by admixture with an equal amount of aniline although a solution of the dye in this solvent exhibited no fluorescence. An alcoholic solution of methylene-violet RRA from which the fluorescence had been removed at the ordinary temperature by the addition of alcoholic chrysoidine recovered its fluorescence to some extent.Solutions of ferric and aluminium hydroxides of protargol and of egg-albumin were not coagulated ; on thawing clear solut.ions were obtained. C. F. B. N. H. J. M. Course of Chemical Reactions at High Temperatures. FRANZ FISCHER (Chem. &it. 1906 30 1291-1295).-The influence of temperature on chemical equilibrium and on reaction velocity is discussed. As an instance of the formation of a substance more stable a t high temperatures than a t low the fact that nitric oxide is obtained by the rapid cooling of a strongly heated mixture of nitrogen and oxygen is quoted The application of similar reasoning to the reaction between oxygen and hydrogen explains why hydrogen peroxide is formed on allowing a hydrogen flame to impinge on a block of ice.Several experiments which have already been published elsewhere are described in illustration of the preparation and properties of ozone and hydrogen peroxide. P. H. Heat Developed on the Addition of Bromine to Certain Unsaturated Substances. WALDIMIR F. LUGININ and IWAN A. KABLUKOFF (.L Chim. Phys. 1906 4 489-506).-The heat developed by the addit.ion of bromine by unsaturated acids of the oleic series and by certain terpenes has been measured the reacting substances being dissolved in carbon tetrachloride. For the addition of 1 molecule of bromine to 1 molecule of the unsaturated compound the followiog amounts of heat were found to be developed undecenoic acid 28,609 cal.; oleic acid 28,757 cal. ; elaidic acid 27,131 cal. ; euricic acid 29,166 cal. ; limonene 28,548 cal. ; carvene 27,545 cal. ; d-pinene (from Russian terebenthene) 35,784 cal. ; I-pinene (from French terebenthene) 35,509 cal. ; pinene 36,237 cal. The differencesGENERAL AND PHYSICAL CHEMISTRY. 73 between the last three numbers are attributed to differences in the purity of the products. By the addition of 2 molecules of bromine to one molecule of limonene 50,420 cal. are developed but this result is somewhat indefinite because about 5% of the bromine used is liberated as hydrogen bromide. In the case of carvene the proportion of hydrogen bromide liberated is smaller (2*9%) and 48,936 cal.were obtained for the heat of addition of 2 molecules of bromine to 1 molecule of the hydrocarbon. The results indicate that the addition of the second molecule of bromine is accompanied by a much smaller development of heat than that of the first. Relation between Viscosity and Absorption Coefflcient for Liquids. MAX TRAUTZ and H. HENNING (Zeit. plqsikal. Chem. 1206 57 25 1-%4).-A summary of the authors' experimental data bearing on Winkler's formula (Abstr. 1892 556) and a discussion of that worker's recent paper (Abstr. 1906 ii 342). Some objection is taken to Winkler's method of calculation and the experimental results do not conform so closely t o the formula as has been represented. A t the same time it is noteworthy that for variations in the absorption coefficient from 0.01 to 1300 and variations in the molecular weight of the absorbed gas from 2 to 160 the value of k remains of the same order of magnitude Direct Measurements of the Osmotic Pressure of Solutions of certain Colloids.BENJAMIN MOORE and HERBERT E. ROAF (Biochem. J. 1906 2 34-73).-The osmometer employed consisted of a metal chamber between the two halves of which a membrane of parchment paper could be fixed and supported by a platinum grid. The one half of the metal chamber was filled with the colloid solution and connected with a mercury manometer the other half contained water occasionally a salt solution. A 10% gelatin solution was found to give a steady osmotic pressure of about 70 mm. of mercury at 30". This pressure persists for a long period and is therefore not an effect due to the initial presence of crystalloids f o r these would gradually diffuse out and the osmotic pressure would fall to zero.The observed osmotic pressure increases with rise of temperature ; the increase however is greater than if the osmotic pressure were proportional to the absolute temperature and is probably connected with the partial dissociation of the solution aggregates in the gelatin solution. When the gelatin solution is kept for a short time a t 70" or SOo and is then brought back to 30° the value of the osmotic pressure is considerably higher than that obtained before the heating; there is however a gradual recurrence to the previous value. If heating a t 80" or 90" is continued for some time a permanent change is effected and the solution has a higher osmotic pressure.A t the same time the physical properties of the gelatin solution are altered and proteid derivatives are formed which pass through the membrane. Other colloids besides gelatin which give a distinct osmotic pressure are the serum proteids and gum acacia. Potato starch and gum tragacanth on the other hand appear to have such a high state of aggregation that they give no indication of osmotic pressure H. M. D. J C. P.74 ABSTRACTS OF CHEMICAL PAPERS. At the stage of hydrolysis of starch at which the blue colour with iodine has just disappeared the dextrine present give a permanent osmotic pressure. Sugars and uric acid behave as true crystalloids and pass through the membrane. Addition of a considerable quantity of magnesium sulphate to serum causes after an initial rise a fall of osmotic prwsure to a value which is less than that given by the original serum.A lecithin or lanolin membrane is permeable to crystalloids and hence the supposed presence of such a membrane will not explain the peculiar content of the cell in crystalloids. It is more probable that the cell protoplasm has selective absorptive powers for different ions and that such ions exist in the cell in combination or adsorption with the cell substance. J. C. P. Velocities of Diffusion of Electrolytes. GIUSEPPE BRUNI and B. L. VANZETTI ( A t t i 3. Accad. Lincei 1906 [v] 15 ii 705-715).- Tha authors have repeated some of the experiments made by Bhscaglioni and Purgotti ( A t t i If. 1 s t . Bot. I’avicc 1905 New Series ll) on the dihfiion of electrolytes and have verified their results.When the two ends of a column of 5% gelatin are placed in contact with solutions of two salts such as silver sulphute and barium chloride capable of forming two different precipitates two distinct septa form in the gelatin one of each precipitate. That this result is t o be explained by the independent migration of the ions as stated by Buscaglioni and Purgotti (Zoc. cit.) is improbable and the authors regard it as highly probable that the formation of the two septa depends on supersat,uration phenomena (compare Morse and Pierce Abstr. 1904 ii 14). The hydrolysis of coloured salts such as copper sulphate may be readily demonstrated by diffusion into gelatin. The velocity of migration of a substance into gelatin is nearly independent of the concentration of the solution.Several cases have been met with which are in disaccord with Buscaglioni and Purgotti’s hypothesis (Zoc. cit.) that the separate ions diffuse with velocities inversely proportional to the square roots of their weights. T. H. P. Modifled van der Waals’ Equation. ALEXIUS BATSCHINSKI (Arm. Physik 1906 [iv] 21 1001-1012).-The author’s form of the equation is p v = K T - A ( l / k - l/v)/(v-h) in which A k and X are constants. The applicability of this equation is tested in the case of ethyl ether with satisfactory results and the conclusion is drawn that for ethyl ether and all other substances which obey the foregoing formula the molecular mass is the same in the liquid as in the gaseous state.J. C. P. Kinetics of Successive Reactions of the First Order. ADAM RAKOWSKI (Zeit. physikaE. Chem. 1906 57 321-340).-A general mathematical study of successive reactions of the first order. “he reactions for which detailed formul~ are deduced are those of the following types (1) ill -+ AL.. -+ ; (2) JIl -3 Mz ;2 J3 ; (3)GENERAL AND PHYSICAL CHEMISTRY. 75 XI -+ M2 -+ M3 -+ M,. Even for these comparatively simple cases and especially in the last case very complicated formulp are obtained and the testing of the equations by experimental data becomes exceedingly difficult if not impossible. J. 0. P. Ionic Velocity and Ionic Hydration I. CHARLES G. CARROLL (Amer. Chem. J. 1906 36 694-509. Compare Jones and Carroll Abstr. 1905 ii 73).-The author shows that the following general law holds for solutions in water or methyl alcohol.The velocity of a given ion under a constant fall of potential is directly proportional to the valency of the ion the cube root of the ionic (atomic) volume and the specific inductive capacity or dielectric constant of the solvent and is inversely proportional to the viscosity coefficient of the solvent. The validity of the law is tested by means both of relative ionic velocities and of conductivity data and it is shown that the law admits of a simple physical interpretation. T. H. P. Ionic Reactions in Acetone. P A u r DUTOIT and HENRI DEMIERRE ( J . Chirn. Phys. 1906 4 565-575).-The velocity of the reaction CH,Cl*CO,H + MI = CH,I*CO,H + MCl where M represents sodium potassium or ammonium has been investigated in acetone solution. The progress of the reaction was followed by measurement of the electrical conductivity which is almost entirely due to the alkali iodide for the chloride is almost insolublo in acetone and the two acids are only very slightly dissociated. The experimental data obtained a t temperatures between 36' and 41' indicate that the reaction is bimolecular.The value of the constant calculated from the equation k = l/t*x/a(a - x) decreases as the reaction progresses and this is found t o be due to the fact that the change is reversible equilibrium being reached when about 94°/0 of the iodide has been converted into chloride. I n order to eliminate the influence of the reverse change in de- ducing the dependence of the velocity on the concentration the initial velocities only have been compared.The value of k thus obtained in a series of experiments increases as the concentration diminishes but kip where p is the molecular conductivity of the solution remains constant. From this the authors conclude that the iodion is the active agent in the substitution process. H. M. D. Relationehip of the Dissociation of Dissolved Substances to their Reactivity. J. TlMMERMANs (Bull. soc. chi??&. Belg. 1906 20 305-31 3).-A critical review is given of recent results experimental apd speculative bearing on Ostwald's generalisation regarding the intimate relation between chemical activity and state of ionisation in the course of which the work recorded in the following abstracts is referred t o Rohland 1899 ii 144 ; 1900 ii 468.Naumann 1899 ii 423; 1904 ii 819 ; 1905 ii 29 30. Briihl 1899 ii 10. Kahlenberg 1899 ii 397; 1902 ii 301. Walden 1901 ii 11; 1904 ii 227; 1906 ii 149. Moisean and Dewnr 1903 ii 419. Plotnikoff 1904 ii 156. Beckmann 1904 ii 235. Franklin and76 ABSTRACTS OF CHEMICAL PAPERS. Eraus 1905 ii 298. Patten 1905 ii 36. Wallace Walker Trans. 1904 85 1082. Mercuric aluminium and ferric chlorides furnish solutions in methyl chloride which exhibit marked electrical conductivities and the solution of the last-named may even be electrolysed yielding iron and chlorine. Solutions of ferric chloride mercuric chloride and cadmium nitrate in methylal do not conduct electricity and similarly mercuric and ferric chlorides dissolved in dimethylamine scarcely conduct whereas aluminium iodide i n dimethylamine shows marked conductivity.Solutions of potassium iodide in melted iodine and of potassium chloride in liquid chlorine do not conduct. Molecular weight determinations by the cryoscopic method of solutions of potassium iodide or mercuric iodide in melted iodine gave high results probably due to the formation of additive compounds in these two cases. These results the author concludes lend no support to the views put forward by Briihl and by Wallace Walker (Zoc. cit.). Shroeder 1905 ii 306. T. A. H. Reactions between Acids and Methyl-Orange. VICTOR H. VELEY (Zeit. physikccl. Chew. 1906 57 147-167).-The action of acids (almost all organic) on a dilute solution of methyl-orange has been studied by a colorimetric method.Each tube of the colorirneter was charged with 20 C.C. of the methyl-orange solution. To the one tube were then added successive portions of 0.1 C.C. of acid (generally N/200) and fresh methyl-orange solution was added to the other until the colours in the two tubes mere equally intense. If II is the number of added portions of acid and y is the increase of depth in cm. in the second tube then for all added acids which obey Ostwald's dilution law the relation between x and y is a linear one and the order of value of the coefficients in the equations y = hlx y = kax &c. is the same as the order of the affinity constants of the acids. If the added acid does not obey Ostwald's dilution law; then the x-y curve represents a para- bola or a straight line not passing through the origin.The co- efficients in the linear equations for some dibasic acids exhibit a slight variation with concentration analogous to the variation with concen- tration i n the case of the afinity constants of these acids. J. C. P. Velocity of Formation of Hydrogen Bromide from its Component Elements. MAX BODENSTEIN and S. C. LIND (Zeit. physikal. Chem. 1906 57 168-192).-At the temperatures at which the authors' experiments were carried out namely 224.7' 251 *4O 277-5" and 301*3O the reaction H + Er2 = 2HBr proceeds completely from left to right. The results for the velocity are readily reproduced and are independent of the glass surface of the containing vessel and of traces of foreign gases. The rate of change is very satisfactorily expressed byithe empirica.1 formula dx/dt = k(cc - x ) ( b - x)*/lm + z/(6 - :)I in which a and 6 are the initial concentrations of hydrogen and bromine respectively x is the hydrogen bromide concentration at time 8 and rn is a constant for which the value 5.0 is chosen.The introduction of a denominator on the right hand side of the foregoing formula is dueGENERAL AND PHYSICAL CHEMISTRY. 77 t o the fact that the combination of hydrogen and bromine is retarded by hydrogen bromide so that the reaction is an example of negative autocatalysis in a homogeneous system. This retardation is a specific property of hydrogen bromide and it is shown not to be due to any chemical combination of that compound with the bromine. The presence of other gaseous substances such as carbon tetrachloride air or water vapour does not cause -any retardation of the reaction although iodine has a very marked effect in that way.It will be noticed from the fore- going formula that the velocity of the change is proportional to the square root of the bromine concentration and hence it is suggested that the bromine reacts in the atomic condition. The temperature coefficient of the reaction has a value which is if anything slightly greater than usual. J. C. P. Rates of the Reaction6 in Solutions containing Potassium Bromate Potaasium Iodide and Hydrochloric Acid. ROBERT H. CLARK (J. Physical Chem. 1906 10 679-700).-The iuiluence of the concentration of each of the reagents on the velocity of the change has been determined by the measurement of the initial velocities.The fractional alteration in the concentration of the reagents was so small in the experiments that the velocity could be treated as practically constant during the interval of change and only in certain cases was it necessary to apply a small correction factor. The reaction was stopped at a particular point by the addition of an excess of an ammonium hydrogen carbonate solution. The data obtained indicate that the rate of liberation of iodine is proportional to the concentrations of the bromate and the iodide and to the square of the concentration of the acid The addition of chlorion in the form of sodium chloride does not alter the velocity which remains the same whether the reaction takes place in air or in an atmosphere of carbon dioxide If iodide and bromide are both present in solution these are oxidised independently the bromide much more slowly than the iodide.In the presence of iodine which forms the ion 18' the velocity of the change is slightly increased in consequence of the oxidation of the triiodion by the bromic acid. This influence is however too small to cause any disturbance in the general character of the change. The value of the temperature coefficient of the velocity between 0' and 30° is 1.85 per rise of 10'. H. M. D. Influence Exerted by a Salt in Various Concentrations on the Velocity of Decolorisation of Aqueoue Solutions of Organic Dyes under the Influence of Light. GUIDO BARGELLINI and ALDO MIELI ( A t f i R. Accad. Lincei 1906 [v] 15 ii 773-778).- From the preliminary experiments here de*cribed it appears that the addition of gradually increasing amounts of a salt to a solution of a dye such as methylene-blue causes firstly a diminution and after- wards an increase in the rate at which the solution is decolorised by the action ( I f light. There is hence a minimum rate of decolorisation corresponding with a detinite concentration of salt.With safranine and potassiuiu chloride decoloration is slowest in a solution containing about 5% of the ,salt and with methyl-violet and magnesium sulphate VOL. XCII. ii. 678 ABSTRACTS OF CHEMICAL PAPERS. the minimum is attained when about 2.5% of the sulphate is present. With solutions of magenta the phenomenon is complicated by the formation of a considerable amount of precipitate.T. H. P. A Substance which Possesses Numerous Liquid Phases of which Three at least are Stable in Regard to the Isotropic Liquid. FRANS M. JAEGER (Proc. K. Akad. Wetensch. Amsterdam 1906 9 359-362. Compare Abstr. 1906 i 742).- Cholesteryl cinnamate the substance in question softens at 151' and thereafter exhibits a brilliant display of colour until a t 157' it is a thick orange-red doubly refracting fluid. If this i s stirred the fluid crystals ar0 seen to form links of lustrous bright green and violet lamina. As the temperature is raised the liquid becomes thinner and at 199.5' it is nearly colourless. At that temperature the mass suddenly assumes a white enamel-like appearance thickens and separates into two anisotropic liquid layers. The interference colours have now entirely disappeared.On further heating the substance changes at 201 03' into a clear colourless isotropic liquid. The author thinks that this is a case where the transitions involved in the change solid -+ liquid occur continuously instead of suddenly. This makes it possible to realise a whole series of labile intermediate conditions. J. C. P. Formation of Hydrosols by the Interaction of Ions. ALWRED LOTTERMOSER (Chem. Centr. 1906 ii 1597 ; from Verh. Ges. deut. Nuturf. Aerxte 1905 ii 87-89. Compare Abstr. 1906 ii 429).-A11 amorphous silver salts may be obtained in the form of hydrosols by the interaction of ions pFovided that an upper limit of concentration is not exceeded. For this purpose it is necessary how- ever that an excess of one of the reacting ions should be present.If the latter have combined completely with the salts which practically are un- dissociated precipitation of the hydrogel takes place. The existence of the hydrosol in the case of a given ion reaction depends therefore on the presence of a definite quantity of anions or silver ions whilst on the other hand no excess of either ion can be detected in the hydrogel. The hydrosols which owe their formation to silver ions and are therefore positively charged are readily gelatinised by OH ions ; bi- and multivalent ions have also a strong action. The hydrosols prepared by means of anions are practically insensitive however to the action of H ions. E. W. W. Hydrate Theory. HARRY C. JONES (Zeit. pl'yeikul. Chem. 1906 57 244-250).-In replying to Biltz's criticism (Abstr.1906 ii 737) the author defends his claim to be the originator of the hydrate theory as applied to abnormal freezing point depression phenomena. J. C. Y. Theory of Colloidal Envelopes ('' Umhiillung "). Ultramicro- scopic Observations. LEONOK MICHAELIS and LUDWIG PINCUSSOHN (Uiochsm. Zeit.,lSCrG 2,251-263. Compare Quincke Abstr. 1902 ii 200 ; Bechhold 1904 ii 650 ; Biltz 1904 ii 392)-In continuationGENERAL AND PEYSICAL CBEMISTRY. 79 of Raehlmann’s work (Phys. Zeit. 1904,4 884 ; PfEuger’s Archiu 1906 112 128)’ experiments have been conducted with suspensions of mastic and indophenol obtained by dissolving the two substances in alcohol pouring into water and mixing the suspensions using the same volume of indophenol but varying amounts of mastic sus- pension.The indophenol suspension itself exhibits a red pseudo- fluorescence (Siedentopf Physill;. Zed. 1905’6 855) but after twenty- four hours the solution has become colourless and a crystalline precipitate of indophenol is obtained. When a small amount of mastic suspension is present a diminution of the pseudofluorescence is observed and in the ultramicroscope numerous white mastic particles are observed with only a few coloured indophenol particles When the amount of mastic has reached a certain valne the solution has a pure blue colour when observed in transmitted light the red pseudo-fluorescence has completely disappeared and also the coloured indophenol particles and the suspension may be kept for seventy-two hours without undergoing alteration. Such a mixture contains the minimum amount of mastic required to ‘‘ protect ” the indophenol present Experiments made on counting the number of particles present in a given volume and also in their electrolytic properties lead t o the view that the two types of particles unite together to form larger particles and that probably the union occurs between a particle of each component. The union produces a change in the physical properties of the particles and this is closely related to the “pro- tective action.” The disappearance of the pseudofluorescence indicates that the union produces an alteration in the inner structure of the indophenol particles.J. J. 8. A Natural System of arranging the Chemical Elements in which they fall into the Periodic Groups based solely on the Atomic Volumes and the Combining Weights. JAMES MONCKMAN (Chem. News 1907 85 6-9).-As an extension of Lothar-Meyer’s observations on the relation existing between the combining weights and the atomic volumes of the elements the author has constructed a series of new curves by plotting combining weights against the tangents of the angles that Meyer’s curves form with the horizontal. The ten curves so obtained arrange themselves naturally into three sets which are termed exionic endionic and central; they comprise respectively the lightest the heavier and the heaviest elements. The central set consists of three lines and contains the triplets iron cobalt and nickel ruthenium rhodium and palladium osmium iridium and platinum. It is claimed for this system of classification that it satisfies a number of qualities that must be found in any perfect system of classification For details of the system as well as a tabular re- presentation the original should be consulted. New Absorption Apparatus for Gases. C. JOSEPH G~LICH (Chem. Zeit. 1906 30 1302).-The liquid absorbent is introduced into the top of a narrow vertical tube resting on a pivot and escapes from it by side-tubes attached at right angles. By the flow of the liquid the tube is sot into ritpid rotation and a fine spray of liquid is P. H. 6-280 ABSTkACTS OF CHEMICAL PAPERS. projected into the surrounding gas thus ensuring efficient absorption. For details of the apparatus the diagram in the original paper should be consulted. P. H.