ii. 81 (3eneral and Physical Chemistry Recalcdlation of Atomic Refractions. FRITZ EISENLOHR (Zeitsch. plqsikal. Chent. 1910 75 585-607).-As certain factors for example optical exaltation mere not sufficiently taken into account in the older tables of atomic refrsctivities the author has recalculated the refraction for the Ha D Hp and Er lines and the dispersions Hp - Ha and Hv - Ha for a large number of non-aromatic compounds and from the results the atomic refractions and dispersions given in the accompanying table have been calculated. The experi- mental data is due mainly to Bruhl and t o Landolt and has been amplified by the author. On the basis of the new data benzene is found to show no optical anomaly. A tomic Re fractions. Ha. D. m* CH group .................. 4.598 4'615 4.668 Carbon ........................2'413 2.418 2.438 Hydrogen .................. 1 -092 1 -1 00 1 '1 15 Oxygen (in CO group) _.. 2'189 '2.211 2.247 (in ethers) ...... 1.639 1'643 1'649 Chlorine .................. 5.933 5.967 6.043 Bromine ..................... 8.803 8.865 8'999 Iodine ...................... 13.757 13.900 14.224 (in OH group) ... 1'522 1.525 1.531 Ethylene linking ......... 1 %86 1.733 1 *824 Acetylene ......... 2 -828 2 '398 2 -506 Hr 4'710 2.466 1.112 8.267 1.662 1.541 6'101 9'152 14.521 1.893 2.538 HB-TIa 0.071 0.025 0.023 0.057 0.012 0 006 0.107 0.211 0 '482 0'135 0.139 Hy-Ha. 0.113 0'056 0.029 0.078 0'019 0-015 0.168 0-340 0.775 0.200 0.171 G. S. Refraction and Magnetic Double Refraction of Solutions of the Rare Earths. G. J.ELIAS (Bey. Deut. physikal. Ges. 1910 12 956-962. Compare Abstr. 1908 ii 549).-The refractive index of an aqueous solution of neodymium nitrate has been measured for wave- lengths in the neighbourhood of the yellow absorption band. The data indicate the occurrence of slightly anomalous dispersion in this region. Preliminary observations of the double refraction produced by a solution of erbium nitrate in a transverse magnetic field are also recorded. The double refraction increases with the intensity of the magnetic field ; it is greater for blue than for yellow light and appears to be somewhat anomalous in the immediate neighbourhood of the absorption bands. No magnetic double refraction could be detected in solutions of neodymium nitrate mttnganous nitrate or ferrio chloride. The double refraction phenomena are examined theoretically.H. M D The Sequence of Chemical Forms in Stellar Spectra. SIR NORMAN LOCKYER (Proc. Roy. Xoc. 1910 A 84 426-432).-A brief account is given of recent work on the classification of stars on the YOL. c. ii cii. 53 ABSTRACTS OF CHEMICAL PAPERS. basis of spectroscopic observations and the influence of temperature on the spectra of ‘‘ chemical forms ” characterised by special molecular or corpuscular groupings. H. M. D. The Emission and Absorption of Luminous Gases from Experiments with Continuous Currents of High Intensity. WJLHELM JURGJOHANN (Zeitsch. wiss. Photochem. 19 10 9 84-1 03 105-1 29 141-1 68).-The intensity of the emission spectra obtained from nitrogen hydrogen carbon monoxide and oxygen when subjected to continuous currents in Wehnelt tubes has been measured for currents of varying intensity.I n the case of nitrogen and hydrogen the absorption was also investigated. The data for nitrogen show that the intensity of the first group of bands is proportional to the current density between 1.4 arid 12.7 amperes per square cm.; on the other hand that of the second group of bands increases more rapidly than the current density or the energy consumption. I n the case of bydrogen the variations of intensity of the spectrum do not appear to be related in any simple way with the current density or the energy consumption and the lines of the first subsidiary series behave differently according to the pressure of the gas in the discharge tube.The intensity of the red and green bands of carbon monoxide increases in a linear manner with the current density that of the blue bands more rapidly. With increasing current intensity the energy maximum is displaced in the direction of smaller wave-lengths. The conclusion drawn from these observations is that tho changes in the energy of the radiation from a luminous gas cannot be accounted for in a simple manner by the changes in the intensity of the current or by the variations in the energy consumption or yet by the corresponding changes in temperature. The Resonance Spectra of Iodine. ROBERT W. WOOD (PhpikuZ. Zeitsch. 1910 11 1195-1196 *).-Resonance spectra similar to those investigated previously in the case of sodium vapour are obtained quite readily when a large evacuated bulb containing a small quantity of iodine is subjected to the influence of white or mono-chromatic light.When. the fluorescence is excited by white light its colour is yellowish- green but a reddish-yellow fluorescence is obtained when the rays from a mercury lamp are used as the exciting agent Spectroscopic examination of the fluorescent light of iodine acted on by the mercury rays has shown that the spectrum consists of a series of i:olated lines separated from one another by intervals of about 65-70 Angstrom units. The investigation of the resonance phenomena can be carried out much more easily i n the case of iodine as compared with sodium for the fluorescence effects are developed by iodine vapour at the ordinary temperature. The Glow Discharge in Rubidium and CEesium Vapours. GEORG GEHLHOFF (Be?..Detct. physikal. Ges. 1910 12 963-969)- The phenomena accompanying electrical discharge through rubidium * and Phal. Uag. 1911 [vi] 21 261-265. H. M. D. N. M. D.GENERAL AND PHYSICAL CHEMISTRY. ii. 83 and cssium vapour have been examined in a similar way to that described previously in the case of sodium and potassium (Abstr. 1910 ii 679). Discharge commences through rubidium vapour at about 120° and through cmium at 60-709 I n both cases the negative glow is at first yellowish-green in colour but a t a higher temperature rubidium exhibits a bluish-violet and caesium a sky- blue colour. The nature of the emission spectra of the negative and positive glow and the effect of the admission of small quantities of hydrogen nitrogen and air are described in detail (ccmpare following abstract).H. M. D. The Emission of the Series and Fundamental Spectra in the Glow Discharge of the Alkali Metal Vapours. GEORG GEALHOFF (Ber. Deut. physikal. Geu. 19 10 12 970-974).-The author compares the glow discharge emission spectra ol sodium potassium rubidium and caesium. Ail four metals show the lines of the principal and subsidiary series in the spectrum of the anodic glow light although i t is probable that the principal lines in the spectra of potassium rubidium and cssium are due t o the surrounding positive sheath. I n the case of sodium the spectrum of this positive column of light contains only the principal series of lines; in the case of the three other metals both principal and subsidiary series are observed although the secondary series is very faint for potassium.The negative glow spectrum of sodium shows the principal and subsidiary series whereas the corresponding spectra for the three other metals exhibit the secondary series lines and the fundamental spectrum. These relationships indicate a close relationship between the spectral emission and the atomic weight of the metal. It is assumed that the principal series lines are due t o impacts between the atoms and electrons of small velocity those of the subsidiary series t o impacts with electrons of intermediate velocity whilst the fundamental spectra are caused by impact of the atoms with high speed electrons. By taking into account the varying electropositive character of the alkali metals it is shown that various phenomena observed in connexion with the glow discharge can be satisfactorily accounted for. When the discharge tube contains both sodium and potassium vapours the positive glow and sheath exhibit the potassium spectrum.The negative glow shows the principal and subsidiary series of sodium lines and in addition the subsidiary series and the fundamental spectrum of potassium. This effect is explained in terms of the above assumption as due to the less electropositive character of the sodium. H. M. D. Absorption in Luminous Hydrogen. RUDOLF LADENBURG (Ber. Deut. physikal. Ges. 1910 12 1018-1022).-Polemical against Pfliiger (compare Abstr. 1910 ii S l l ) . H. M. 0. The Ultra-violet Absorption of Benzene. L. GREBE (Zeitsch.wiss. Photochem. 1910 9 130-140).-The absorption of ultra-violet rays by benzene in liquid and vapour form has been investigated. The spark discharge between aluminium electrodes under water was 6-2ii. 84 ABSTRACTS OF CHEMICAL PAPERS. used as a source of ultra-violet radiation and absorption spectra were obtained by means of a quartz spectrograph. I n the case of benzene vapour observations were made at temperatures ranging from 17O to 185O and the records show that this variation of temperature is practically without influence on the position of the absorption bands. The same result follows from the observations on liquid benzene at ZOO 43O and 70°. On comparing the positions of the liquid and vapour bauds it is found that the former are displaced in the direction of the visible spectrum by a nearly constant amount represented by 17 AngstrBm units.Prom experiments in ethyl-alcoholic solutions containing 0.1 and 50% of benzene respectively it is found that with diminishing concen- tration the bands are displaced in the direction of smaller wave- lengths. I n a 50% solution in ethyl ether the bands occupy the same position as in the corresponding alcoholic solution. Molecular Vibrations of Solid Substances. A. STEIN (Physiknl. Zeitsch. 1910 11 1209-1212).-0n the assumption that the mole- cules of solid substances are in vibratory movement and that the amplitudes of the vibrations are proportional to the distance between the centres of the molecules in the case of substances which have the same melting point the author deduces an expression for the relative frequencies of the molecular vibrations. This may be written in the form n,/n,= Jzz $</ Jml Vq in which rn n and v denote respectively the frequency the molecular weight and the molecular volume of a solid substance.According to the theory that the absorption of ultra-red rays is a consequence of molecular (or atomic) vibrations the values of n / r ~ ~ can be obtained from observations on the behaviour of solid substances towards ultra-red rays. From the mean frequencies of vibration of the residual rays obtained by Rubens by selective reflexion from plates of rock-salt sylvine potassium bromide and potassium iodide the values of nlln are obtained for the different pairs. On comparison of these ratios with the corresponding JG 2q/ JGl vc ratios it is found that the two series of numbers are approximately the same.A similar agreement is found by a comparison of the data for quartz and fluorspar. The values of Jm Jv are also tabulated for various series of metals having approximately the same melting points. These values indicate the existence of simple relationships for the relative magnitudes correspond very closely with simple integral ratios Influence of Substituent Groups on the Spectrum of Progressive Phosphorescence. JOSEPH DE KOWALSKI and J. DE DZIEEZBICKI (Compt. rend. 191 1 152 83-86. Compare this vol. ii 3) .-The progressive phosphorescence of the following substances has been examined at about - 190' in alcoholic solution benzoic acid the toluic acids the amino- and hy droxy-benzoic acids benzonitrile and the three toluonitriles.The wave-lengths of the principal bmdR are given. The substihent groups are found to have a marked influence on the spectrum of progressive phosphorescence ; those like H. IT. D. _ - - - H. M. D.GENERAL AND PHYSICAL CHEMISTRY. ii. 85 the CO,H and CN groups which extend it towards the red are termed 6‘ bathophosphic,” whilst those which prolong the spectrum towards the violet areecalled hypsophosphic. Examples of the latter are the hydroxy- and methoxy-groups ; the methyl group occupies an inter- Ultra-microscopic Observations. J. AMANN (Zeitsch. Chem. Inn. Kolloide 1913 8 11-15. Compare Abstr. 1910 ii 844),-1f strongly illuminated ultra-microscopic particles are examined by means of an ultra-microscope which is slightly out of focus the image of a particle presents the appearance of a series of concentric light and dark rings showing interference colours.With this arrangement the sparkling of the ultra-microscopic particles is much more evident than when the ultra-microscope is in focus. The sparkling effect is attributed to a variation in the amount of light reflected from a particle at different times in consequence of the irregular shape of the particle and its rotational motion. These movements are termed krypto-kinetic in order to distinguish them from Brownian motion. The phenomenon in question has also been observed in sections of yiixrtz from different sources. Solutions of potassium ferro- and ferri-cyanide under the ultra- microscope exhibit numerous large particleq. When subjected to white light LL large number of small particles suddenly appear and this so-called photo-phase is supposed to be the immediate cause of the evolution of oxygen from hydrogen peroxide when a solution of this substance is brought into contact with a previously insolated ferro- or ferri-cyanide solution (compare Weigert Abstr.1908 ii 5 ) . Observations of the colour of fluorescein under different conditions are recorded. According to these it appears that a colloidal substance may exhibit four totally different colours according to the method of illumination. H M. D. mediate position between these. w. 0. w. Micro-polarisation. EMIL FISCHER (Bey. 191 I 44 129-1 32. Compare Donau Abstr. 1908 ii 647).-The following pieces of apparatus are described 1.A small glass cylinder 20 x 7 mm. provided with a glass stopper. This is used for the preparation of small amounts of solutions of known concentration. 2. A pyknometer of the usual form but with very thick walls. so that it holds about 0.0’7 C.C. of liquid. These are of white glass 50 x 1.5 mm. and do not hold more than 1 C.C. 4. A narrow glass tube drawn out to a capillary for transferring the liquids from one vessel t o another. The light must be strong ; either incandescent gas light or the apparatus made by Schmidt and Haensch for obtaining homogeneous light from the Nernst lamp is recommended. An accurate balance is essential and very volatile solvent7 are to be avoided With tubes narrower than those described the sedimentation of the solutions does not take place. Action of Ultraviolet Rays in Accelerating Chenii cal Reactions and in Modifying a State of False Equilibrium.JEAN POUGNET (J. Pharm. Chem. 1910 [vii] 11 540-543).-Ultrs- 3. Observation tubes for the polarimeter. 3. J. S.ii. 86 ABSTRACTS OF CHEMICAL PAPERS. violet light accelerates the rate of change of plastic into rhombic sulphur of vitreous selenium into crystalline selenium and of barley sugar into the crystalline form. It also increases the velocity (I) of the transformation of potassium manganate into potassium permangan- ate ; (2) of the reaction between potassium permnnganate and oxnlic acid (3) of the inversionof sucrose by acids and (4) of the saponification of ethyl acetate by sodium hydroxide. T.S. P. Principal Types of Photolysis of Organic Compounds by Ultra-violet Light. DANIEL BERTHELOT and HENRY GAUDECFION (Compt. vend. 1910 151 1349-1352. Compare Abstr. 1910 i 349 814).-Hydrogen predominates in the gases formed when alcohols of the type CH,R*OH undergo decomposition through exposure to ultra- violet light ; the hydrocarbon R is evolved with the lower terms of the series with the higher members it remains dissolved in the slcohol. Hydrocarbons of the type R or RR’ are not formed in the case of alcohols with a branched chain but owing to the greater ease with which the lateral chain is broken methane and ethane occur in con- siderable quantity. Photolysis occurs more readily with the lower members of the alcohols aldehydes and acids than with the higher ones.Hydrogen (2 vols.) and carbon monoxide (1 vol.) are evolved from ally1 alcohol Carbon monoxide is the principal gaseous product when aldehydes are decomposed the proportion rising as the series is ascended ; hydro- carbons are also formed. Acraldehyde undergoes rapid polymerisation under the action of the light; the gas evolved contains carbon monoxide (80 vols.) carbon dioxide ( 5 vols.) hydrogen (5 vols.) and an un- saturated hydrocarbon (10 vols.). Citral gives carbon monoxide and hydrogen. Camphor forms carbon monoxide (5 vols.) and methane Acids give chiefly carbon dioxide accompanied by hydrogen and carbon monoxide with hydrocarbons in the earlier members of the series the latter diminishing as the series is ascended. (1 vol.). W. 0. W. Action of Light on Chlorophyll.P. A. DANGEARD (Compt. rend. 1910,151,1386-1388).-A simple method of demonstrating the effect of light of different wave-lengths on chlorophyll or other substances is to mix an alcoholic solution of chlorophyll with collodion and coat a glass plate with the mixture. When dry the plate is exposed to a good spectrum. After some hours colourless bands appear where action has taken place; the first band to appear corresponds with the principal absorption band of chlorophyll. w. 0. w. Canal Rays. JORANN KOENIGSBERGER and K. KILCHLING (Ber. Beut. pfiysikal. Ges. 1910 12 995-1017).-0bservntions relating to the nature and properties of canal rays are described with reference to the phenomena of dissociation neutralisation diminution in velocity and deflexion.H. M. D. HORACE H. POOLE (Phil. Mag. 1911 [vi] 21,58-62. Compare Abstr. 1910 The Rate of Development of Heat by Pitchblende.GENERAL AND PHYSICAL CHEMISTRY. ii. 57 ii 176).-New determinations of the rate of heat development from pitchblende by the former method gave as the results 8.15 6.5 and 6.2 ( x cal. per hour per gram) the last experiment which was the best agreeing well with the former determination 6.1. On the assumption that 1 gram of radium generates 110 cal. per hour the theoretical development should be 4.4. F. S. The Ratios which the Amounts of Substances in Radio- active Equilibrium Bear to One Another. HUGH MITCHELL (Phil. Mag. 1911 [vi] 21 40-42).-From the general solution of the equation giving the quantity of the nth substance in a dis- integration series after any time t the parent substance being initially free from products the correct relation between the quantities of the members of a radioactive series in equilibrium is deduced.The ratio of the number of atoms of the lzth substance to that of i t and all the preceding substances in equilibrium with it is the same as the ratio of the average life of the &,h substance to that of the parent sub- stance. The correct relation is thus A& = xn/(xI + x2 + x3 + . . . . xn) instead of A& = x,/xl as usually written. The correct relation makes no assnmption as to the relative periods of the parent and its products beyond the requirement that the parent element should be the longest- lived of the series. F. s. The Relation between Viscosity and Atomic Weight for the Inert Gases; with its Application to the Case of the Radium Emanation.ALEXANDER 0. RANKINE (PhiZ. Mag. 1911 [vi] 21 45-53).-From the temperature-coefficient of the viscosity of the inert gases Sutherland's constant C is calculated and from it the viscosities at the critical temperatures are deduced and found to be proportional to the square-root of the atomic weights for argon krypton and xenon. The value so found for helium is about sixteen times the value so calculated from its atomic weight but direct experiments by Schmitt of the viscosities of argon and helium from - 193' to + 184' have shown that whereas argon conforms to helium departs from Suther- land's equation the more seriously the lower the temperature. It is probable tbe equation fails so near to the absolute zero.The unknown critical temperature of neon calculated from the assumption that its viscosity at the critical temperature is proportional to the atomic weight and from Sutherland's equation is 6 1.l'A. Calculated from the relation that Sutherland's constant is proportional to the critical temperature it is 62.7"A. The viscosity of radium emanation a t Oo calculated from its atomic weight critical temperature and the value of Sutherland's constant which is taken as proportional to the latter is 2.13 x The molecular volume calculated from this is 16-97 times that of helium and larger than that of any other gas in the group. The deduction that the critical temperature and therefore Sutherland's constant is pro- portional to the fourth power of the true atomic radius is found to hold good strictly for all the gases except helium for which a departure is to be expected.F. S.ii. 88 ABSTRACTS OF CHEMICAL PAPERS Researches on Helium. IV. Absorption of Helium by Salts and Minerals. ARNALDO PIUTTI (Rend. Accad. Sci. Ris. Nut. Napoli 1910 [iii] 16 253-255 *).-The author finds that helium is absorbed by certain fused salts and minerals and points out the bearing which this has on the calculations of the age of rocks which have been made on the assumption that the helium contained in them has all been derived from radioactive elements. Crystals of potassium sulphate deposited in the course of a month from an aqueous solution in an atmosphere of pure helium did not absorb the gas.On the other hand antimonite borax and boric anhydride after fusion in an atmosphere of helium show the spectrum of the gas when they have been powdered in the air. I f air is passed through fused borax or boric anhydride and the fused mass then suddenly cooled by means of liquid air the solid obtained shows the spectrum of helium. No helium can be found when air has not been passed through the fused salt. The 7-Rays of Thorium and Actinium ALEXANDER S. RUSSELL and FREDERICK SODDY (Phil. Nag. 1910 [vi] 21 130-154. Com- pare Abstr. 1909 ii 460 851; 1910 ii 474).-The two types of thorium y-rays from mesothorium-2 and thorium-D resemble that of radium-C both in the ratio of the intensity of the y- to that of the /I-rays and in their penetrating power. Thorium-B gives the most penetrating y-rays known the absorption-coefficient X being from 8 to 21% less than for radium-C y-rays whereas for the mesothorium y-rays X is from 4 to 25% greater than for radium according to the experimental disposition used for the measurements.The y/P ratio for mesothorium is from 1-0 to 0.8 and for thorium-D 0.69 to 0.51 times that for radium-C. The proportion of y-rays contributed by the two thorium products in equilibrium in minerals is very similar. The y / P ratio of actinium is only from one-eighth to one-sixteenth of that of radium-C so that aotinium in this respect resembles uranium-X. The y-rays of actinium are abnormally highly absorbed by lead the absorption curve showing two sudden changes in the value of X at 0.3 cm. and 0.85 cm.whereas for zinc and aluminium the curves are exponential X being about 1.9 times that for radium-C. For lead the ratio ranges from about 8 times to about 2.2 times according to the part of the range examined and the disposition employed. Admixture of mesothorium with radium can be detected by the departure of the y-ray absorption curve in lead from the exponential form the y-rays of mesothorium being distinctly the less penetrating. Some generalisations with iegard to y-rays show that they are more allied in properties to the a-rays which precede and follow them in the series than to the /3-rays which accompany them. The penetrabilities of the y-rays from uranium-X mesothorium-2 radium-C and thorium-l) increase as the period of the product diminishes as for a-rays in general whilst the a-rays preceding and following these types of y-rays show the same increase in penetrating power as the y-rays themselves The greater the penetrating power of the yrays the less abnormal the absorption by lead as compared with other metals and the less is the hardening produced by passage through lead.* and Le Radzuczu 1911 8 13-14. R. V. S. F. S.GENERAL AND PHYSICAL CHEMISTRY. ii. 89 The Ionisation of the Atmosphere Due to Radioactive Matter. A. S. EVE (Phil. Mug. 1911 [vi] 21 26-40).-The radio- active theory of the ionisation of the atmosphere is generally satis- factory but some results remain unexplained. Assuming the presence of the emanation in equilibrium with 80 x 10-l2 gram of radium per cubic metre of the atmosphere on the average the ionisation so caused by the a-rays is 1.63 ions (per C.C.per second) while thorium possibly contributes not more than one additional. From Wright’s values of tbe natural ionisation in an aluminium electroscope at Toronto on land and on the frozen surface of Lake Ontario the ionisation contributed by the y-rays from the radium in the earth isestimated to be in the free air about 2.5 which is smaller than that found in many experiments with lead-screened electroscopes and suggests that part of the penetrating radiation comes from the atmosphere. It is calculated however that the penetrating rags of the atmosphere from the emanation and its products must produce an ionisation negligible in comparison with that due to the a-rays and are one twenty-third as intense as the penetrating rays from the earth.The total ionisation of the air from all sources assuming the penetrating rays of the radium and thorium series in the earth to be equal and neglecting those from uranium and actinium is 4.35 which is some- what greater than the value normally found in c l e y weather. The fact that the ionisation over the ocean is scarcely less than over the land is the greatest objection to the radioactive theory of atmo- spheric ionisation for the amount of radium in sea-water is extremely small. It is calculated that there should be a rapid diminution of the effect due to y-rays from the earth with altitude detectable at an elevation of 100 metres while at 1000 metres the effect should be negligible I n an appendix Wulf’s results (PftysikaZ.Zeitsch. 1910 11 811) on the Eiffel tower are referred to as establishing this rapid diminution. F. S. The Radioactivity of the Leinster Granite. ARNOLD L. FLETCHER (Phil. Mag. 1911 [vi] 20 102-lll).-The whole mass of the Leinster granite covering 600 square miles contains radium the mean content being 1.7( x 10-12 gram per gram). Separate specimens showed quantities varying from 0.41 to 4-36 the specimens showing the highest and lowest amount coming from the same locality. Probably the distribution is fairly homogeneous throughout. The mica in the granite constituting about 20:k of the whole Contained about half the radium but the small quantity of thorium present did not seem to be concentrated in any constituent. One crystal of biotite showing countfess pleochroic halos and considerable radioactive darkening contained 11.87 units of radium but no thorium and in the three specimens of granite containing the most radium there was no thorium.I n ten out of twenty-three specimens the ratio between the radium and thorium was between 2.0 and 2.6 x 10-7 the mean ratio for the whole series being 2.4 x 10-7. Details are given of the methods employed including the latest form assumed by Joly’s arrangement for estimating thorium in minerals F. S.ii. 90 ABSTRACTS OF CHEMICAL PAPERS. A Spectroscopic Investigation of the Nature of the Carriers of Positive Electricity from Heated Aluminium Phosphate. FRANK HORTON (Proc. Roy. Xoc. 1910 A 84 433-449).-The value found for elm for the positive ions from heated aluminium phosphate indicates a molecular weight of the carrier of about 28 and the possibility that the molecules are either nitrogen or carbon monoxide.Two hours' heating of a platinum strip coated with aluminium phosphate generated sufficient ions t o allow of a spectroscopic examination of their nature. The spectrum was found to be the same whether the strip was heated simply or in an electric feld under conditions such that the ions would be formed and in the electrodeless ring discharge showed the lines of carbon and oxygen together with others due to mercury and hydrogen whilst in an electrodeless spectrum tube the banded spectrum of carbon monoxide was obtained. The conclusion is arrived a t that the positive ions are molecules of carbon monoxide. F. s.Relationship between Chemical Aenity and the Photo- electric Effects of Potaseium in its Compounds. ROBEBT POHL and P. PRTNGSHEIM (Ber. Deut. physikal. Ges. 1910 12 1039-1048. Compare Abstr. 1910 ii 923).-The photo-electric properties of potassium alloys have been investigated and it is shown that these are connected with the' electro-chemical properties of the metal with which the potassium is associated. A comparison of the results obtained for bismuth antimony and phosphorus alloys shows that the normal photo-electric effect of potassium is displaced in the direction of smaller wave-lengths as the metal in combination with the potassium becomes more electro-negative. The same behaviour is exhibited by the alloys of the series bismuth lead thallium mercury and gold.I n those cases in which a selective photo-electric effect can be observed this is also found to he displaced in the same direction as the electro-negative character of the second component of the alloy becomes more pronounced. A t the same time the range of wave- lengths which give rise to photo-electric activity becomes more restricted. Since the frequency of the active rays affords a measure of the velocity with which the electrons leave the molecule of the photo- electrically active substance it appears that the affinity between a metal and its electrons can be determined from observations on the wave-lengths of the raps which give rise to the photo-electric effect. H. M. D. Transference Experiments with Thallous Sulphate and Lead Nitrate. K. GEORGE FALK (J.Amer. Chem. Soc. 1910,' 32 1555-1571).-A study of the change of the transference number of tri-ionic salts with the concentration was carried out by Noyes (Abstr. 1901 ii 143) with the object of ascertaining whether intermediate ions such as KSO,' in the case of potassium sulphate and NO,Ba' in that of barium nitrate exist in appreciable quantities in solutions of such salts and the results were recorded for potassium sulphate barium chloride and barium nitrate.GENERAL AND PHYSICAL CHEMISTRY. ii. 91 Transference experiments have now been made with 0.03~V- and O'lNsolutions of thallous sulphate and lead nitrate at 25' with the aid of special apparatus. The cation transference numbers found for the thallous salt are 0.479 at 0.03Xand 0.476 at O*liV and those for the lead salt 0.487 at both concentrations.The limit of error in these results does not exceed & 0.003. The transference numbers at zero concentration as calculated from Kohlrausch's extrapolated values of t h e equivalent conductivities of the separate ions are 0.489 for thallous sulphate and 0.503 for lead nitrate and ale thus 2.5-3.2% higher than the values obtained by direct measurement a t a concentration of O.U3-0*1N. It is not certain whether this change of the traiisference number with the concentration is real. If so it might be due in the case of thallous sulphate to the presence of the intermediate ion TISO,'. I n the cace of lead nitrate however the presence of the intermediate ion N0,Ph' would probably cause a change in the opposite direction.The results of these experiments and those of Noyes (lac. cit.) on the transference values of tri-ionic salts indicate that either the inter- mediate ion is not formed to any considerable extent by the dissocia- tion of such salts up to concentrations of 0.1 or O-SN or that if it is formed its equivalent conductivity has a definite value of such magnitude as to render the transference number independent of its concentration. E. G. Method for Making Two Substances React in the Electric Arc. PAUL SABATIER (Compt. rend. 1910 151 1328).-An apparatus similar to that described by Salmon (this vol. ii 15) was employed by the author in 1899 (CongrBs Assoc. francaise I 229). W. 0. W. A New Property of the Magnetic Molecule. PIERRE WEISS (Compt. rend. 1911 152 79-81).-The niagnetic susceptibility of magnetite has been measured at different temperatures between 550' and 900° and the results plotted in the form of a curve.The conclu- sion is drawn that at certain temperatures the magnetic moment of the molecule increases by a definite aliquot portion of the magnetic moment at low temperatures and the supposition is advanced that this is due to a variation in the size of the molecule or in tha distance between the poles. w. 0. w. Use of the Magnetic Field as a m e a n s of Determining Constitution in Organic Chemistry. V. PAUL PASCAL (Bull. Xoc. chim. 1911 [iv] 9 6-12. Compare Abstr. 1910 ii 100 179).- Further evidence of the additive character of magnetic susceptibility in organic compounds is given (compare Abstr. 1910 ii 580) and the values of B (the correction which peculisrities of structure necessitate in calculating molecular suFcepti bility) are given for a number of compounds free from oxygen and chlorine.The atomic susceptibilities for the usual constituents of organic compounds are as C1= - 209.5 x 10-7 ; Br = - 319.2 x 10-7 ; I = - 465.0 x 10-7. Those for nitrogen sulphur and variously linked oxygen have been given follows H = - 30.5 10-7; c= - 62.5 10-7 ; F = - 63.0 x 10-7;ii. 92 ABSTRACTS OF CHEMICAL PAPERS. already (Abstr. 1910 ii 100). For saturated open chain hydrocarbons the value of B is nil and the molecular svsceptibility is given by the formula XM= - 10-7[n62.5 + (2n+ 2)30.5]. For a single open chain ethylenic linking B= + 57 x 10-7 and for two or more such linkings +110 x 10-7.For a double linking between two nitrogen atoms B= + 19 x 10-7 for one between a carbon and a nitrogen atom it becomes +85 x 10-7 and for two such linkings i t is + 106 x 10-7 but for a triple linking as in *CN it falls to + 8 x 10-7. The influence of the cyclopropane nucleus is B= + 75 x which is much greater than that +31 x 10-7 due to the cydohexane nucleus. Similarly for the piperidine nucleus B = + 37 x I n cydohexene derivatives B = + 72 x 10-7 and in cyclohexadiene compounds it is + 110 x 10-7. Theso cases illustrate the general ruIe that the diamagnetism of the molecule in closed chain compounds falls with increase in the number of certain kinds of double linkings present. Solutions. 111. and IV. F. SCHWERS (Zeitsch. phpsikal.Chem. 1910 75 615-620 621-627; Bull. SOC. cAim. 1910 [iv] 7 1072-1077 1077-1083.+ Carnpare Abstr. 1910 ii. 1039).- 111.-Eelationship between the Density and the Magnetic Rotation of the Plane of Polarisation of Binary Mixtures.-It is shown from the data of Sir W. H. Perkin that the ratio between the alterations of the deneity (difference between observed density and that calculated according to the mixture rule) and the alterations of the magnetic rotation (difference between observed and calculated rotation) is a constant for mixtures of the two components in any proportion. The constant 2 has a definite value for each system and when solutions of the same substance (water) in the different members of an organic series are examined 2 is found to increase with increasing molecular volume.It is shown by reference to mixtures of sulphuric acid and water and of nitric acid and water that when electrolytic dissociation occurs 2 diminishes steadily with increasing dilution. IT.-Density Befractivity and Alagnetic Rotation of Dissociating Mixtzcres.-As in the case of sulphuric and nitric acids the constants 2 for hydrochloric hydrobromic and hydriodic acids in aqueous solution also diminish on dilution. When the molecular volume is increased by substitution of iodine or bromine for chlorine the rotation constant diminishes and the refraction constant (calculated in a similar way) increases. When however the increase of molecular volume is due to an increase in the number of atoms in the molecule the effect on the coostsnt,s is the converse of that mentioned above.Electrolytic dissociation brings about an increase of the magnetic rotation of the plane of polarisation and a diminution of the refractivity with reference to the density if the latter is regarded as constant. G. S. The Isolation of Long-waved Heat Rays by Quartz Lenses. HEINRICH RUBENS and ROBERT W. ~ ' O O D (Sitxungsber. K. &ad. Wiss. Berlin 1910 1122-1137).-A method is described by means of which heat rays of wave-lengths ranging from 80p to 2 0 0 ~ Ic and BdI. Acnd. roy. Belg. 1910 850-883. T. A. H.GENERAL AND PHYSICAL CHEMI9TRY. ii. 93 Can be isolated from the rays emitted by an incandescent mantle. The rays from the mantle are passed through two quartz lenses which are suitably placed with reference to the source of light and the isolation of the longwaved rays is essentially dependent on the selective refraction of the quartz lenses The bundle of rays obtained by this method is far from homogeneous but the energy content is much greater than that of the homogeneous rays obtained by selective reflexion.By means of a radio-micrometer the distribution of energy amongst the component rays of the bundle has been examined and these measurements show a maximum intensity in the neighbourhood Observations are also recorded which show the extent t o which the long-waved rays are absorbed by various solid liquid and gaseous substances and a comparison is made between the amounts of these rays and of those obtained by selective reflexion from potassium bromide (A = SZp) which are reflected; by various substances.I n the case of rock-salt fluorspar and glass the proportion of the reflected quartz rays is in good agreement with that calculated from the dielectric constants of those substances. of A = loop. H. M. D. A Simple Automatic Stirrer for Use with the Depressi- meter. LODEWYK TH. REICHEH. (Chem. Weekblad 1910 7 1085-1087). -The author has devised an automatic stirrer for use with the depressimeter. A copper disk supported by a horizontal axle held by a clamp is connected with a jointed iron rod by means of an eccentric attachment. A horizontal arm connects the rod with the vertical stirrer. The bearing of the rod can be moved nearer to or further from the centre of the disk thus diminishing or increasing its eccentricity. The motive power is supplied by a hot air engine through a belt to the copper disk.A. J. W. Vapour Pressures of Binary Mixtures in the Light of van der Waals’ Theory. 11. PHILIPP KOHNSTAMM (Zeitsch. plqsikal. Chem. 1910 75 527-551. Compare Abstr. 1901 ii 145).-Considerable light is thrown on the vapour-pressure curves of binary mixtures by a mathematical treatment based on the introduc- tion by van der Waals of a new factor the vapour pressure of an ideal binary mixture the ‘‘ mixture assumed as uniform,” defined as follows. In the ordinary pu-diagram for a single substance the part of the isothermal joining the points on the diagram representing the gas and liquid phases is a straight line but in the case of a binary mixture the corresponding part of the diagram will be a sloping and not a horizontal line (as evaporation does not occur at constant pressure).The ‘‘ uniform mixture ” is an ideal mixture such that the line in question is horizontal as for a simple substance. By making use of the (hypothetical) critical pressures and temperatures of such mixtures a number of deductions as to the possible types of binary vapour-pressure curves are made. For the method used and the forms of the curves on different assumptions as to the relative magnitude of the factors concerned the original paper shotild bb consulted. G. s.ii ABSTRACTS OF CHEMICAL PAPERS Method for Determining the Molecular Weights of Dissolved Substances by Measurement of Lowering of Vapour Pressure. ALAN W. C. MENZIES (J. Amer. Chern. Soc. 1910 32 1615-1624).-Apparatus is described for the determina- tion of molecular weights in cases in which great accuracy is not required.It consists of an outer tube in which the solvent is boiled attached to a small reflux condenser and an inner test-tube provided with a pressure gauge-tube and a glass stopper. The side-tube from the outer tube to the condenser is of 1 cm. diameter so that the solvent may boil under the barometric pressure and is connected to the condenser by means of rubber tubing furnished with a screw-clip. When the stopper is removed atid the clip closed the vapour of the liquid boiling i n the outer tube escapes through the gauge-tube which is open at both ends into the test-tube. A narrow U-shaped side- tube connects the condenser with the lower part of the outer tube and the condensed liquid is thus returned to the outer tube without cooling the upper part of the test-tube. Both the test-tube and gauge-tube are graduated.The lower end of the gauge-tube is in the form of a bulb which is perforated by a number of holes t o distribute the vapour as it issues. The boiling solvent in the outer tube maintains at a constant temperature the test-tube which contains the solution the vapour pressure of which is being determined. The pressure is measured by the difference in the level of the liquids in the gauge-tube and test-tube. The values of K,,,=p-p' (where p is the vapour pressure of the solvent and p' that of the solution) that is the lowering of vapour pressure in mm. of boiling solvent that would be caused by the presence of 1 gram-mol.of a non-volatile solute in 1 litre of solution have been calculated for various solvents. The molecular weight can be calculated from the formula M = 1000 WKB/L Y x 760 where W is the weight of solute added L the measured lowering of the vapour pressure in mm. V the volume of the solution in c.c. and B the height of the barometer. A series of determinations carried out with this apparatus are quoted and the method is compared with ebullioscopic methods. It is sliown that within the limits of error uf observation the vapour pressures are identical whether determined by static or dynamic methods and that if any difference at all exists betKeen the vapour pressure at the b. p. as measured statically and dynamically of water and alcohol such difference does not exceed 0.01 mm.of mercury. E. G. Convenient Form of Apparatus f o r the Measurement 0 the Vapour Deneities of Easily Volatile Substances. ALAN w. c. MENzIES (J. Amer. Chern. soc. 1910 32 1624-1628).-The apparatus described is a slight modification of that recommended by the author for the determiriationof molecular weights of dissolved substances (preceding abstract). The test-tube is closed a t the lower end by the introduction of 6-8 C.C. of mercury which also serves by its rise in the gauge-tube to indicate the increase of pressure in the test-tube. T ~ Q substance the vayour density of which is to be determined isGENERAL AND PHYSICAL CHEMISTRY. ii. 95 placed in small glass bulbs each provided with a capillary which is sealed off before the final weighing.The bulb is attached to the lower end of the stopper of the test-tube by inserting the point of the capillary into a hole drilled in the stopper and wedging it in with a few shreds of dry asbestos. The hole is made in an oblique direction so that the neck of the bulb may be broken by rotating the stopper in such a way as to force the bulb against the top of the gauge-tube. I n order to obtain a constant for the apparatus a quantity of benzene sufficient to give a rise in the mercury level of about 100 mm. is weighed in a small bulb and the latter is attached to the stopper of the test-tube. A liquid of suitable b. p. is placed in the outer tube and caused to boil steadily so that its vapour surrounds the whole of the closed portion of the test-tube.As soon as the mercury in the gauge-tube has attained a constant level the stopper of the test-tube is rotated in order to break the neck of the small bulb which falls on to the surface of the mercury. The mercury instantly rises in the gauge-tube and reaches a constant level in a few seconds The rise in the mercury column is observed and a correction made by adding the amount of the slight fall of the level of mercury in the test-tube. Prom the data (1) weight of benzene taken (2) molecular weight of benzene and (3) rise of mercury in mm. the required constant K is calculated representing the rise in mm. which would be caused by the volatilisation of one gram-mol. of any substance in the apparatus a t the particular tem pera ture. The molecular weight of a substance can be calculated by means of the formula M - V K / R where li is the rise of the mercury in mm.and W the weight of substance taken. It is not suitable for temperatures above 200° but determinations are quoted which show that the method is both rapid and accurate at other temperatures. E. G. The apparatus is compact and can be easily manipulated. Fall of Temperature in Vapours of High Molecular Complexity at Small Pressures. C. VON RECEIENBERG (Zeitsch. physikal. Chem. 19 10 75 628-636).-Polemical against Hansen (Abstr. 1910 ii 827). No new facts are adduced. G. S. Critical Phenomena in Solutions under the Kardioid Ultra-microscope. W. G. VON LEPKOWSKI (Zeitsch. physikul. Chem. 19 10 '75 608-6 14).-The ultramicroscope in its latest improved form has been employed to elucidate the nature of the opalescence in binary mixtures in the critical region.The results are on the whole in favour of Donnan's theory (Abstr. 1904 ii 240) that the oFalescence is due to minute drops. A mixture of amylene and aniline critical temperature 14*%-14.5° was used for the experiments. On cooling progressively from room temperature ZOO a grey field with a clear circle in the middle was observed; at a certain point the previously clear field appeared as a rapidly agitated mass in which separate particles could not be dis- tinguished. As the temperature is further lowered (but still above the critical temperature) the flickering becomes still more pronounced theii. 96 ABSTRACTS OF CHENICAL PAPER% ultramicroscopic particles moving with enormous velocity.Suddenly relatively large drops appear and the critical temperature is reached. The phenomena appear in the converse order on warming. Particles from the small drops are projected into the bulk of the liquid with enormous velocity ; the drops become smaller and smaller and finally disappear. The remarkable observation was made that when a relatively large drop is warmed until it has disappeared and even the flickering has stopped i t reappears on cooling at the same point and with the same contours. This phenomenon occurs repeatedly on warming and cooling. It may be taken as showing that the rate of diffusion in the critical region is very small. G. S. Sublimation Apparatus. EMIL DIEPOLDER (Chem. Zeit. 191 1 35 $).-The apparatus consists of a wide glass tube sealed at the lower end and closed at the upper by a rubber cork carrying two tubes one of which serves for the entrance of air or other gas whilst tho other is attached to a wider tube which fits accurately into the outer closed tube and acts as a receptacle for the sublimate.The substance to be sublimed is contained in a small beaker placed at the bottom of the outer tube. The apparatus may be used for sublimation either at the ordinary or under diminished pressure. F. E. Therniochemical Studies of Some Binary Compounds of the Alkali and Alkali-earth Metals. ROBERT DE FORCRAND (Compt. rend. 1911 152 27-31).-The author has collated in tabular form the best existing data for the heats of dissolution and formation of the chlorides bromides iodides fluorides and oxides of calcium strontium barium lithium sodium potassium rubidium and cesium. Deter- minations of the heats of dissolution of the following anhydrous compounds were made to complete the series.The values are in Calories. Ca12 + 28.12 Lip - 1.04 RbCl - 4.50 RbBr - 5.96 RbI - 6 50 RbF + 5-80 CsCl - 4.68 CsBr - 6-73 CsI - 8.25 CsP + 8-37. The data are discussed from the point of view of Berthelot’s principle of maximum work. w. 0. w. Thermochemistry of the Silicates. EMIL DITTLER (Zeitsch. anorg. Chem. 1911 80 273-304. Compare Abstr. 1909 ii 970).- The difficulties of obtaining correct cooling and heating curves in the case of silicates owing to the suspended transformation effects are fully discussed and the precautions which must be taken are enumerated.Thermal and optical measurements on diopside from Zermatt adularia from St. Gotthard oligoclase-albite from Wilmington (Delaware) labradorite from Kiew an artificial labradorite artificial anorthite and anorthite from Pizmeda are given. Owing to the extremely small velocity of melting the absorption of heat a t the melting point is not the chief factor in the heating curve so that with different velocities of heating the melting point will occur at diBerent points in the curve. Melting point determinations on large fragments must be avoided since they are more readily super- heated than a fine powder and thus give higher results.GENERAL AND PHYSICAL CHEMISTRY. ii. 97 Geologically freezing points are more important than melting points and are generally lower than the latter owing to undercooling.The temperatures a t which cry stallisation begins to take place in various minerals are as follows artificial diopsicle 1?90-1250° ; artificial labradorite 1 230° ; labradorite from Kiew 1200-1 160' ; artificial anorthite 1310' ; anclrthite from Pizmeda 1200'. The temperatures a t which crystallisation is complete could not be determined. T. S. P. Determinations of the Law of Chemical Attractions between A t o m s from Physical Data. RICHARD D. KLEEMAN (Phil. Mag. 19 11 [vi] 21 83-102).-From data of Ramsay and Young and Mills relating to surface tension and latent heat of benzene chlorobenzene carbon tetrachloride and methyl formate and to the critical density of ethyl ether Py-dimethylbutane Pedimethylhexane isopentane n-pentane rb-hexane n-hept ane n-octane benzene cyclohexane fluoro- benzene chlorobenzene bromobenzene iodobenzene carbon tetra- chloride and stannic chloride various consequences of the mathe- matical expressions deduced in previous ptLpers are tested.F. 8. Recovery of Hammered Materials. L ~ ~ o N GUILLET (Compt. rend. 1910 151 1127-1128).-Experiments on hard and soft steel and on nickel show that complete recovery on annealing characterised by attainment of maximum length occurs in each case at 750-800". The curves showing elasticity on traction exhibit two elastic limits one constant and characteristic for the metal in the ordinary state the other variable with the temperature of recovery. The metals in this respect behave as heterogeneous materials of which only one constituent shows the effects of hammering w 0.w. Some Abnormal Temperature-coefflcients of the Molecular Surface Energy of Organic Substances. PAUL WALDEN (Zeitsch physikccl Chern. 1910 75 555-577).-According to the EotvGs- Rarnsny-Shields rule the temperature-coefficient of the molecular surface energy d [ y ( Mv)3]/dt is about 2.121 for unimolecular sub- stances for associated substances it is less and for dissociated sub- stances greater The author now finds that the coefficient in question is 6-21-5.35 for tristenrin and 5.57-4.92 for tripalmitin between 60' and 120° whereas the molecular weights according to cryoscopic determinations with a number of organic solvents are normal. The coefficient for the isobutyl ester of ricinoleic acid amounts to 3.24-3-30 between 2 3 O and 86' whilst its molecular weight is normal.The above rule therefore does not hold even approximately and the deviations appear to be the greatest for complex substances of high boiling point. The surface-tension of solutions of tripalmitin and tristearin in benzene has also been measured. The average value of the coefficient for both solutions calculated according to the mixture rule is about 2-18; the solutions therefore behave as non-associated and non-dissociated liquid systems. On the other hand two hydroxyl-free solvents with abnormally small coefficients of molecular surface energy succinonitrile and formamide VOL. C. ii. 7ii. 98 ABSTRACTS O F CHEMICAL PAPERS. have been investigated. From the results of surface-tension mc'asure- ments the coefficient for the nitrile between 80' and 100' is 0.565 indicating a moleciilar complexity of 7.27 from 100-118' it is 0.601 the complexity being 6.63.For formamide between 5' and 60° the coefficient varies from 0.638 to 0.710 corresponding with a molecular coniplexity of 6975-5.1 7. The conclusion that these two compounds are highly associated is confirmed by cryoscopic and ebullioscopic measurements. G. S. EMIL HATSCHEK (Zeitsch. Cliem. Ind. Kolloide 1911 8 34-39. Compare this vol. ii 19)-On the basis of certain assumptions relating to the form and distribution of the disperse phase in the dispersive medium the author has examined the changes which should result from the disturbance of the normal arrangement of the disperse phase as a consequence of the relative movement of juxtaposed layers of an emulsion and has deduced a relationship between the viscosity of the emulsion and that of the dispersive medium.The effect of this relative movement such as is involved in the flow of the emulsion through capillary tubes is dependent on the relative magnitudes of the surface tension and the viscosity of the disperse phase if the proportion of this is very large and a decrease in the viscosity of the disperse phase is shown to be accompanied in certain circumstances by an increase in the viscosity of the emulsion. For a disperse system of this type there is a critical velocity of displacement of juxtaposed layers a t which an abrupt change takes place in the viscosity and this velocity is determined by the rate at which the particles of the disperse phase regain their original structural form when deformed as a result of the lateral displacement of these layers.For velocities of displacement greater than this critical value the relationship between the viscosity of the emulsion and that of the dispersive medium is expressed by ql = 77 If/( v'- l) in which yr and q are the viscosities of the emulsion and the dispersive medium and f i s the ratio of the volume of the emulsion to that of the disperse phase. For velocities of displacement smaller than the critical value the mathematical treatment of the question presents difficulties but it is shown that the viscosity of the system under these conditions must be greater than for greater velocities The Viscosity of Dispersoids.H. M. D. The Phenomenon of Agglomeration in Finely Powdered Substances. M. C. SCHUYTEN (Chem. Zeit. 1910 34 1357).- A sample of pure iron powder kept for fifteen years in a well stoppered bottle which had not undergone oxidation had formed a globular mass with a peculiar goose-neck shaped excrescence. Barium sulphate prepared by mixing equivalent proportions of solutions of barium chloride and ammonium sulphate gradually formed conical excrescence reaching a length of 6-7 cm. The author ascribes these phenomena to pbysical actions. Globular Appearance of Certain Precipitates. FRIEDBICH BARDACH (Chem. Zeit. 191 1 35 33).-The globular or flocculent L. DEE.GENERAL AND PHYSICAL CHEMISTRY. ii. Y9 appearance assumed by finely divided substances such as precipitated barium sulphate etc.has been fully discussed by Quincke (Abstr. 1902 ii 200) and others who attribute the phenomenon to the action of osmotic pressure surface tension etc. The outer surface of the flocks has a characteristic appearance whilst the interior frequently exhibits a cell-like structure. w. P. s. Parallelism between the Occurrence of Cohesion and the conditions Necessary for Dyeing. AC’GUSTE ROSENSTIEHL (Bull. Xoc. chim. 1911 [iv] 9 12-18. Compare Abstr. 1909 ii 796).- A number of cases are quoted showing that the conditions under which cohesion occurs between the same or different substances are exactly analogous to those under which the dyeing of fabrics takes place and it is suggested that cohesion between dye and fabric plays the principal rAle in this operation although chemical affinity between dye and fabric may also be operative Deduction of the General Phase Theorem for Adsorption Systems. P.N. PAWLOPF (Zeitsch. physilal. Chem. 1910 75 552-554).-The form of the phase rule for a system capable of adsorption is deduced mathematically; it is similar t o that for a system in which adsorption does not occur. Condensed disperse systems capable of absorption have six kinds of variables whereas T A. H. colloidal adsorption systems are characterised by seven variables. G. S . A New Proof of the Existence of an Adsorption Equili- brium in Hydrosols. ALFRED LOTTERMOSER and P. MAFFIA (Ber. 1910 43 3613-3618).-Experimental results are recorded in support of the view that electrolytes are adsorbed by hydrosols in accordance W i t h an exponential law.Freshly precipitated ferric hydroxide was dissolved in ferric chloride solution and n portion of the electrolyte removed by filtration through collodion. The contents of the filter were then diluted with water and the solu- tion again filtered this process being repeated until a hydrosol containing relatively little chlorine was obtained. From this hydrosol a series of others were prepared by addition of varying quantities of water and each of these was then subjected to filtration through collodion. The chloride in the various filtrates was estimated and by difference the adsorbed chloride was ascertained. Denoting by c the number of millimols. of chlorine per C.C. of the filtrate and by xlm the millimols.of chlorineadsorbed by one gram of ferric oxide it is found that the experimental data can be satisfactorily represented by the equation x/m = P.cl/P in which /3 = 2-83 and l / p = 0.163 are constants. Similar results were obtained in the investigation of a ferric oxide hydrosol prepared by addition of ammonium carbonate to a solution of ferric chloride. ‘‘ Anomalous ” Adsorption. WILLIAM M. BAYLISS (Zeitsch. Chena. Id. Kolloide 191 1 8 2-4).-The so-called ‘‘ anomalous ” adsorption recently observed by Biltz and Steiner (Abstr. 1910 ii 830) can be accounted for on the basis of the author’s observations on electric H. M. D. 7-2ii. 100 ABSTRACTS OF CHEMICAL PAPERS. adQorption ” (Abstr. 1906 ii 344). I n these experiments filter- paper was the adsorbent and this becomes negatively charged in contact with water.I n consequence of this electric charge electro- positive colloids are attracted by the paper whilst electro-negative colloids are repelled. In presence of neutral salts the negative charge is ‘partly or completely neutralised by the cation and as a result the adsorption of electro-negative colloids is facilitated whilst that of electro-positive colloids is hindered. This neutral salt action ex- plains the absence of ‘‘ anomalous ” adsorption in the experiments of Biltz and Steiner when dialysed night-blue was used instead of the commercial substance. It also accounts for the approximately normal adsorption observed in the experiments in which dilute solutions of the dye were examined. H. M.D. Capillary Analysis of Colloidal Solutions. NAIMA SAHLBOM (XoZZ. Chem. BeiThefte 19 10 2 79-141).-The behaviour of colloidal solutions of it number of inorganic substances has been investigated by observations of the capillary efYects produced when long strips of filter paper are supported with the lower edges immersed in the solutions. It has been found that the positive colloids are precipitated at the surface of contact whereas negative colloids rise through the capillary network without hindrance. The precipitation of tha positive colloids is not due to chemical action but to the setting up of an E.M.F. as a consequenco of the streaming of the liquid through the capillaries ; this potential difference brings about the discharge OF the positively charged colloidal particles.If the capillaries are not sufficiently narrow the resulting E.M.P. is not large enough to give rise to complete pre- cipitation and a rise of the positive colloid is observed. A detailed examination of the phenomena observed in the case of colloidal ferric hydroxide has shown that the free acid present in an incompletely dialysed solution prevents complete precipitation. Addition of acid to a completely dialysed hydrosol has the same effect. The capillary phenomena exhibited by solutions of colouring matters are in many cases closely similar to those observed with inorganic colloids. EL M. D. Capillary Electrical Precipitation of Positive Colloids. FRITZ FICHTER (Zeitsch. Chem. Ind. Kolloide 1911 8 1-2. Compare preceding abstract).-The precipitation of positive colloids as a result of the production of electrical forces by difeusion through capillary tubes has been further examined by experiments with parallel glass plates.When a drop of a dialysed colloidal solution of ferric hydroxide is placed on one of the plates and the second plate is allowed to sink slowly on to the former i t is found that the thin layer of solution between the plates exhibits a sharply defined circular opalescent zone in consequence of the precipitation of the positive colloid. A similar phenomenon is observed with colloidal solutions of chromic oxide zirconium dioxide and thorium dioxide. With coloured solutions of basic dyes such as night-blue and methylene-blue the visible effect is still more pronounced. Prom the dimensions of the circular zone of precipitation and theGENERAL AND PHYSICAL CHEMISTRY.ii. 101 weight of the solution between the plates the critical distance between the plates requisite for the production of the precipitation can be calculated. In the case of ferric hydroxide this was found to be 0.0322 but this is very much smaller than the value (0,075 mm.) expected on the basis of experiments with capillary tubes. The much smaller value is probably due to mechanical displacement of the precipitation zone when the plates are brought together. In similar experiments with solutions of negative inorganic colloids and acid dyes no precipitation effects are observed. H. M. D. Origin of Internal Pressure in Solutions. V. POLOWZOFF (Zeitsch. physikal. Chem. 1910 75 513-526).-According t o Tammann (Innere Xrafte und Eigenschuften der Losungen; Leipzig 1907) the greater internal pressuro in a solution as compared with the pure solvent is due to attraction between solvent and solute and is of the same nature in electrolytes and non-electrolytes.Drude and Nernst (Zeitsch. physikcd. Chem. 1894 15 79) on the other hand ascribe the increased pressure at least in the case of electrolytes t o the effect of the electrostatic field of the ions (theory of electrostriction) and in accordance with this view it has been found that the complete ionisation of a mol. of a binary salt is attended by a contraction (resulting from the increase of pressure) of 13-14 C.C. In the present paper further evidence is advanced in favour of the electrg- striction theory.From observations on the density of solutions of monochloroacetic acid in different dilutions the degree of dissociation of the salt being known it is calculated that the complete ionisation of a mol. of the salt is attended by a contraction of about 10 C.C. On the other hand the apparent molecular volume of ethyl alcohol in benzene only varies slightly from dilute to very concentrated solutions although the molecular complexity greatly diminishes on dilution. This is difficult to reconcile with Tammann’s theory as an increase in the number of molecules should cause increased attraction between solvent and solute and therefore increased internal pressure and consequent contraction The data of Ritzel (Abstr. 1907 ii 740) on the compressibility of solutions of ethyl alcohol in benzene are also in favour of the view that the internal pressure does not increase on dilution.Finally it is shown by an indirect method that the dissociation of amyl acetate into amylene and acetic acid is attended by an increase and not by a diminution of volume. This is in agreement with Nernst’s theory-as no ions are concerned there can be no electrostriction. The electrostriction theory cannot of course account for the increased internal pressure caused by dissolution of non-electrolytes such as the sugars and glycerol in water ; these solutions require further investigation. G. S. Influence of Foreign Substances Dissolved in the Mother Liquor on the Faces of Crystals of Meconic Acid and on their Pseudopolychroism. PAUL GAUBERT (Compt.rend. 1910 151 1134-1 136. Compare hbstr. 1906 ii 152).-Meconic acid occurs in OrthorhQmbic hemimorphic cry staia having curved faces andii. 102 ABSTRACTS OF CHEMICAL PAPERS. showing marked double refraction [a b c = 0.613 1 0.900 ; a = 63O p= 138’1. Addition of different substances to the solution in which the crystals are growing may modify the crystalline form to a con- siderable degree ; thus picric acid produces elongation in the direction of the axis a whilst methylene-blue elongates the axis 6. The extent of elongation depends on the amount of foreign matter absorbed by the crystal. A large quantity of methylsne-blue produces hemimorphism the crystals being H-shaped. Addition of carbamide to the mother liquor causes the crystals to exhibit pseudopolychroism when viewed through a Nicols’ prism.This appears to be due to the presence of microscopic bubbles which produce partial absorption through repeated reflexion. w. 0. w. Theory of the Phenomena of Transition between Colloidal and True Solutions. P. P. VON WEIMARN (Zeitsch. Chem. Ind. Kolloide 191 1 8 24 -33).-A theoretical paper in which the author discusses the relationships be tween unsaturated supersaturated and colloidal solutions in reference to the action of the dispersive and coagulating forces. The hydrosols of slightly soluble substances are qualitatively differentiated from the corresponding unsaturated solutions by the fact that continuous condensation takes place in the former. I n accordance with this view the disperse particles are not amorphous but are characterised by vectorial structure.The forces which determine the vectorical arrangemept are peculiar to the molecules and the play of these forces is not inhibited by the excessive development of the surface which is characteristic of colloidal systems. Observations relating to the precipitatioa of supersaturated solutions of barium sulphate are cited in support of the author’s views. H. M. D. The Micellary or ‘‘ Colloidal ” State. GIOVANNI MALFITANO (Koll. Clhern. Beihrzfte 1910 2 142-212).-A theoretical paper in which the relationsbips and differences between the colloidal and crystalloidal states and between the processes of coagulation (floccu- lation) and crystallisation are examined. In part the theoretical con- siderations have reference to experimental observations publisbed previously (Abstr.1909 ii 473) on the behaviour of solutions of ferric chloride. H. M. D. The Saturation Capacity of Colloidal Compounds. WALTH~RE SPRING (Bull. SOC. chim. Belg. 1910 24 446-453 ; Arch. Sci. plt,ys. 32at. 1910 [iv] 30 561-571).-0bservations relating to the influence of saponin on suspensions of soot in water are described which seem to show that the capacity of colloidal substances to form complex aggregates is limited in a similar way to that observed in the case of chemical combination. From preliminary experiments with solutions of saponin it appears that this substance occupies a position intermediate to the colloids and crystalloids. Its solutions show the Tyndall effect and under theGENERAL AND PHYSICAL CHEMISTRY.ii. 103 influence of a current the saponin moves towards the anode ; on the other hand dialysis experiments show that it undergoes diffusion. To ascertain the influence of saponin on the durability of soot suspensions solutions containing from 7.5 to 0.0045% of saponin were shaken up with the same quantity of soot carefully freed from fatty matter by extraction with benzene and the times required for the clarification of the solutions were observed. From these data it appears that R solution containing between 3.5 and 3% of saponin is most favourable to the persistence of the suspension. If the solutions are made either slightly acid or alkaline clarification takes place in all cases very quickly and there is no optimum effect. When a current is passed through the most durable suspension the soot migrates to the anode although a pure aqueous soot suspension shows the opposite effect.This indicates the formation of colloidal aggre- gates as a result of the adsorption of saponin by the soot. The effect of saponin on the behaviour of suspensions of soot when subjected to filtration has been examined and it appears that the adsorption of the soot by other substances such as filter-paper is to a large extent prevented by reason of the formation of soot-saponin aggregates. The view that the formation of colloidal aggregates or adsorption compounds is limited by a definite saturation capacity is supported by this and a number of other observations of similar character which are described in detail.H. M. D. Colloidal Chemistry and the Phase Rule. W. P. A. JONKER (Zeitsch. Chenz. Ind. Kolloide 191 1 8 15-18).-The applicability of the phase rule to colloidal systems is discussed and the conclusion drawn that the relatively large amount of surface exhibited by colloidal substances is without influence on the nature of the equilibrium relationships from the phase rule point of view. The coaguIation of colloids is not comparable with the separation of a liquid into two layers for the former process is not accompanied by by an increase in the number of phases. Objection is taken to the introduction of surface tension as a new independent variable in the consideration of colloidal systems for surface tension is quite different in character from temperature and pressure in that its magnitude cannot be varied at will.I n reference to the connexion between the amorphous and crystal- line states it is stated that realgar is a suitable substance for experimental investigation. This can be obtained in amorphous or crystalline form by melting together equivalent quantities of sulphur and arsenic. The crystalline variety melts at 320° whereas the amorphous becomes liquid at about 130° and when heated to 350' its properties change in a continuous manner. These observations are cited as evidence in support of the view that the amorphous and crystalline forms are to be regarded as distinct phases. IT. M. D. The Equilibrium between Potassium Hydrogen Carbonate and Trihydrated Magnesium Cttrbonate. NANTY (Compt. wnd. 1910 151 1352-1354).-Polernical against Eogel (Abbtr.1886,ii. 103 ABSTRACTS OF CHEMICAL PAPERS. 121) and Buchner (Zeitsch. Elektrochern. 1908 14 63). has re-examined the conditions of equilibrium in the reaction and is unable to confirm the existence of two distinct limits of formation and decomposition for the reverse reactions at 20-30”. For the same temperature at constant pressure the actions appear to The author KHCO + MgCO,,SH,O + H,O MgC03,KHC0,,4H,0 converge towards a common limit. w. 0. w. Electron Conception of Valency. I(. GEORGE FALK and JOHN M. NELSON (J. Amer. Chena. Xoc. 1910 32 1637-1654).- Thomson’s hypothesis that each linking between the atorus of a compound is due to the transference of a negatively-charged corpuscle from one atom to the other is considered in its application to various classes of organic compounds and references to its application to inorganic compounds are quoted from the papers of Noyes (Abstr.1908 ii 349) and Ramsay (Trans. 1908,93 774). It is shown that all cases of isomerism. connected with the presence of a double bond whether between like or unlike atoms can be interpreted on the basis of the direction of the valencies (that is the direction of transference of the corpuscles) .without recourbe to spatial configurations. The theory of directive valencies is also employed to account for the existence of certain isomerides and to account for certain reactions which have not been explained previously. E. G. Stokes’s Forniula. MAX REINGANUM (Bey. Deut. pliysikal. Ges. 1910 12 1025-1038).-1n view of the frequent application of Stokes’s formula in the calculation of the diameters of small particles from observations on their velocities of movement under the influence of EG known force the author has examined the conditions under which this formula holds good.For very small particles such as the ions in gases the formula is not valid and it is shown that the determination of the dependence of the velocity of such particles on the pressure of the gas may be used as a means of ascertaining whether Stokes’s equation is applicable or not. In the use of the small metallic particles investigated by Ehrenhaft (Physikal. Zeitsch. 1910 10 318 940)) the uncorrected equation is inapplicable and the calculated diameters of the smallest particles are in consequence incorrect.On the assumption that the small particles can be regarded as elastic spheres similar to gas molecules the author deduces a new formula connecting the radii of the spheres and their velocity of movement. According to this the radius of the particles is propor- tional to the velocity whereas in Stokes’s formula the radius is proportional to the square-root of the velocity. For a given radius the new formula shows that the velocity is inversely proportional to the density of the gaseous medium but according to the older formula the density of the gas is without influence on the velocity of the particles. From the data of Ehrenhaft the dimensions of the particles calculated from the two formulm are compared. I n the case of gold and platinum the calculated values are nearly the same for the largerGENERAL AND PHYSICAL CHEMISTRY.ii. 105 particles but the author’s formula gives much smaller values for the radii when the most slowly moving particles are considered. H. M. D. Automatic Rapid Mercury Pump for High Vacua. A. BEUTELL (Chem. Zeit. l910,34,1343-1343).-With one of the author’s mercury pumps a vacuum of 0*0007 mm. can be obtained in ten minutes and one of 0.0001 mm. in twenty minutes whilst after twenty-five minutes it was impossible to measure the vacuum with a McLeod gauge. A photograph of the pump as mounted is given but no details. A Modified Separating Funnel and Washer for Heavy Liquids. HARFORD M. ATKINSON (Chem. News 1910 102 30S).- The funnel consists of an ordinary tap funnel with a second tap and short tube inserted about half-way up the side of the bulb.The Only 0.29 kg. of mercury is necessary to work the pump. T. s. P. washing liquid etc. can be washings performed without running off the heavy liquid. The drying with calcium chloride can also be carried out in the funnel. N. C. run off by this side tap and further Filtering Apparatus for Substances which are Hygroscopic or Altered by Exposure to Air. WIL- HELM STEINKOPF (Chem. Zeit. 1910 34 1358).-On the suction flask A is placed a Buchner funnel B having a ground flange the edge of which is turned upwards to prevent the lid from slipping ; the funnel is closed by means of a ground-glass lid C furaished with a neck. Into this fits the adapter B which has a tube E bent a t right angles and a side tubulus J’ into which the neck of the reaction flask G is ground.This neck is bent a t such an angle that when the bulb of the flask is turned upwards the contents liquid and pre- cipitate readily flow into D. The neck of the flask G also has a tubulus H which during the reaction proper may be connected with a reflux condenser or a gas-delivery tube but during filtering is fitted with a separating funnel containing the washing liquid. H is placed not in the-position shown in the diagram but so much to one side thatii. 106 ABSTRACTS OF CHEMICAL PAPERS. the contents will not run into it when the flask is rotated in F. To prevent moisture entering during the filtration phosphoric oxide tubes are attached t o E and to the side-tube of A leading to the pump.As the solvents used are generally ether benzene light petroleum and the like i t is advisable to place on the dry filter paper a cut and ground inverted funnel J which keeps the paper in position and also prevents any splashing of the precipitate on to the sides G€ the Biichner funnel If desired an inert gas can be transmitted through E. L. DE I(. Automatic Extraction of Aqueous Liquids by Organic Solvents of Lower Density. RICHARD KEMPF (Chem. Zeit. 1910 34 1365-l366).-The apparatus consists of a flask surmounted by two tubes placed concentrically to one another ; the vapour from the boiling ether etc. passes through the annular space between the tubes to a condenser. The condensed ether drops into a funnel tube which is placed concentrically in and reaches to the bottom of the inner tube which is closed below and contains the liquid to be extracted. A flat glass spiral is twisted round the stem of the funnel like an ascending stair and so forces the ether into intimate contact with the liquid to be extracted as it ascends through it. The ether collects on tbe surface of the aqueous liquid and flows through two openings in side of the inner tube back into the flask below. L. DE I(. Apparatus for Preserving and Measuring Poisonous Hygroscopic or Low-boiling LiqUids.wrLHELM STEINKOPF (Chem. Zeit. 19 10 34 13 19).-The liquid under investigation for example anhydrous hydrogen cyanide is directly distilled into the cylindrical vessel in which it is to be preserved. Sealed into this vessel is a narrow tube reaching down to the bottom and connected with a measuring tube through a stopcock. Another tube containing a stopcock is also sealed into the top of the vessel and by connecting this tube with a source of pressure the liquid in the vessel may be forced over into the measuring tube the various stopcocks being appropriately manipulated. The whole apparatus in made of glass. T. S. P. Aluminium Apparatus for Use in the Laboratory. HUGO MASTBAUM (Chem. Zeit. 1910 34 1319).-Condensing spirals made of aluminium tubing are very serviceable in distillations occuring in wine and brandy analyses. Boilers of the same material may also be used in the production of steam for distillations. T. S. P. A New Apparatus for the Preparation of Liquid and Solid Air for Demonstrations. MAX BAMBERGER (Oesterr. Chem. Zeit. 1910 [ii] 13 137-138).-The paper contains a sketch with description of an apparatus suitable for the production of small quantities of liquid air. It consists of two series of copper tubes and a steel flask ; the apparatus is cooled and the enclosed air subjected to a pressure of 120 atmospheres. F. M. G. M.