ii. 165 General and Physical Chemistry. Atomic Refraction of Phosphorus in Certain Organic Phos- phorus Compounde. A. E. ARBUZOV and A. A. IVANOV ( J . Russ. Y h y s . Chem. SOC. 1915 47 2015-2027).-The results obtained by Zecchini (A. 1893 ii 353 354; 1894 ii 221) and by Kovalevski ( J . Russ. Phys. Chem. SOC. 1897 29 217) are discussed. I n the authors’ measurements the index of refraction was determined a t 20° f o r the D Ha H and H hies and the values of M(n2- 1)/ DT ( ~ 2 + 2) calculated. The values of the atomic refraction of phos- phorus are given for the D line (1) according t o Conradi; with the three compounds containing quinquevalent phosphorus the values are calculated also on the assumption t h a t all the oxygen is singly linked to the phosphorus; (2) according to Traube.The results are as follows Atomic refraction of phosphorus (D). M(n2- 1 ) <-A- -. Compound. Line. D( n2+ 2) Conradi. Traube. 42-78 42.57 43.33 P( OEt) I”H 7.45 7.12 ......... \g 43.79 D 40.67 For 0’ 40.50 5.34 41.08 For 0” 41-43 4-57 3 1.75 For 0’ 31.61 5.62 32-08 For 0” 32.35 4-86 41.90 For 0‘ 41-74 5.05 42-32 For 0’’ 42.66 4.28 5-01 ... ki POEt(OEt) D 5.47 ... POH(OEt) D 4.64 ...... PO(OEt) For the atomic refraction of phosphorus f o r the D line calculated according t o the Lorenz and Lorentz formula Kovalevski (Zoc. cit.) found from PCl,*OMe 7-72 ; PCl,*OEt 8-22 ; PCl,*OPr 8.01 ; PCl,*O*C,H 8-04 ; and PCl,*O*C,H, 7.60. The atomic refraction calculated from Zecchini’s results f o r a compound regarded by this author as P(OEt) is 2.97 ; according to Arbuzov’s results (A.1907 i 8 174 275) this compound is probably impure diethylphos- phorous acid OH*P(OEt) and recalculation of Zecchini’s result o n this assumption gives the value 4.96 f o r the atomic refraction this number agreeing well with the author’s value 4.86. F o r the four compounds of the table given above t h e authors have determined the values of D;‘’. the molecular volume and the atomic volume of phosphorus according to Kopp’s data; the last of these magnitudes has the values 32-25 26.55 27.38 and 32.79 for P(OEt) POEt(OEt) POH(OEt) and PO(OEt) respectively VOL. cx. ii. 8ii. 166 ASSTtlACTS OF CHEMICAL YAP EKS. For the same four compounds the ratio of the molecular volume to the molecular refraction f o r the D line has the values 4.77 4-88 4.89 and 5-07 the mean being 4.9.Further the ratios of (1) the molecular volumes and (2) the molecular refractions are fox P(OEt),/POEt~(OEt) (1) 1.03 (2) 1-Q5 ; f o r P(OEt),/ YOH(Oht) (1) 1.31 (2) 1-35 ; and for P(OEt),/YO(OEt) (1) 0.96 (2) 1.02. The ratios 1.03 1.05 1-31 1.35 and 0.96 1.02 have the approximately equal values 0.98 0.97 and 0.94. T. H. P. The Neutral Atom and the Poeitive Ion as Carriers of the Band Spdctrum and of the Series Spectrum of Hydrogen. J. SIraRIc ( A ) z n . Physik 1916 [iv] 49 179-200).-The conditions under which the many-lined (band) and the series spectra of hydrogen are emitted lead to the view t h a t the series spectrum is emitted by positively charged hydsogen ions and the band spectrum by neutral hydrogen atoms.The fact t h a t the lines of the band spectrum show no appreciable canal ray Doppler effect affords evi- dence t h a t high-speed neutral atoms are not present in any measur- able quantity in hydrogen canal rays. It is probable that these lines are emitted as a result of collisions with the canal says. H. M. 1). The Single-line Spectra of Magnesium and otner Metals and their Ionising Potentials. J. C. MCLENNAN ( J . E’~~uddin Inst. 1916 181 191-207.* Compare A 1915 ii 657).-Previous experi- ments have shown t h a t the vapours of mercury cadmium and zinc give rise t o spectra which consist of a single line when tha heated vapours are bombarded by electrons the energy of which lies within certain well-defined limits. It has now been found t h a t magnesium vapour behaves similarly the wave-length of the singleline emission being h 2852.22.The wave-lengths of the single-lines in the emission spectra of mercury cadmium and zinc have been found to be identical with those of bands in the corresponding absorption spectra and this re’lation is also shown by magnesium the absorption spectrum of which contains bands atl ~2852.22 and h2073.36. Since these lines are the first members of Paschen’s combination series v = 2,p2 - m,S and v = 1.5,s - m,Y respectively it. follows t h a t the absorption spectrum of magnesium is analogous to the spectra of the vapours of mercury cadmium and zinc. On the assumption t h a t this analogy extends to calcium stron- tium and barium and thallium it is calculated t h a t the single-line emission spectra which would be emitted by these vapours when suitably bombarded by electrons should have the wave-lengths h 4226.91 h 4607.52 h 5535.69 and h 5350.65 respectively.According to Franck and Hertz (Ber. Deut. physikal. Ges. 1914 ‘11 512) the8 single-line frequency of mercury is identical with t h a t calculated from the minimum ionising potential on the basis of the quantum theory. By the same method the author has calculated the minimum ionising potentials of all the metals referred to above. ’ and Proc. Roy. SOC. 1916 [A] 92 305-312.GENERAL AND PHYSICAL CHEMISTRY. ii. 167 It is considered however that this method of correlating single-line frequencies with minimum ionising potentials is difficult to reconcile with Bohr’s theory. A Comparison of the Arc and Spark Spectra of Nickel Produced under Pressure.E. G. BILHAM (Phil. Nag. 1916 [vij 31 163-170).-According to Duffield (A 1915 ii 658) the dis- placements per atmosphere of reversed nickel lines are greater for the arc spectrum than for the spark spectrum. I n view of the fact that the pressure of 10 atmospheres which were employed in the author’s observations on the spark spectrum coincides with a region in which Duffield’s observations on the arc spectrum show ail abnormally high rate of shift per atmosphere it has been con- sidered advisable to subjectl the data to a more rigorous comparison. The conclusion reached is t h a t the behaviour of lines easily reversed o,r tending towards reversal is approximately the same for both arc and spark liims.Differences are shown on the other hand by unreversed lines. This result is exactly the opposite of that reached by Duffield in his analysis of the data. The High Frequency Spectra of the Elements from Gold to Uranium. MANNE SIEGBAHN and EINAR FRIMAN (PhysikuZ. Zeitsch. 1916 17 17-18 *).-Measurements have been made of the wave- lengths of the stronger a-line in the X-ray spectra of the1 elements gold mercury thallium lead bismuth thorium and uranium. The relation between the atomic number and the square root of the frequency is found to be linear as required by Moseley’s formula if the atomic numbers assigned to thorium and uranium are 90 and 92 respectively. H. M. D. Absorption Spectra of the Vapours of Inorganic Salts. E. J. EVANS (Phil. Mag. 1916 [vi] 31 55-62).-The examination of the absorption spectra of the vapours of ammonium chloride mercuric and inercurous chloride cadmium chloride bromide and iodide has shown thatl the vapours have no well-defined absorption lines or bands in the region h 2500 t o h 6700.With the possible exception of ammonium chloride the salts show general selective absorption in the ultraviolet which in the case of the cadmium salts is greater for the iodide and bromide than f o r the chloride. H. M. D. H. M. D. H. M. I>. The Rotation Dispersion of Hydrogen. PAUL SCIIERREH (Physil~al. Zeitsch. 1916 17 18-21).-A theoretical paper in which i t is shown that the observed rotation of the plane of polar- isation exhibited by hydrogen in a magnetic field can be calculated on the assumption that the changes prolducd in the interior of the atom by the application of the magnetic field are in accordance with the requirements of Newtonian mechanics.The Magnetic Rotation of the Plane of Polarisation in Titanium Tetrachloride. 11. L. H. SIERTSEMA (Proc. K. AkacZ. Wetensch. d msterdarn 1916 18 925-932. Compare A. 1915 ii H. M. D. * and Phil. Mag. 1916 [vi] 31 403-406. 8-3ii. 168 ABSTRACTS OF CHEMICAL PAPERS. 610).-Thes experimental observations on the magnetic rotatory dispersion of titanium tetrachloride1 are1 shown to be in agreement with thel requirements of Lorentz’s theory. The Life of Radium. ELLEN GLEDITSCH (Amer. J. Xcz. 1916 [iv] 41 112-124).-Boltwood’s method of finding the period of radium by separating the whole of the ionium present in a uranium mineral and determining the rate of grolwth of the radium from i t in terms of the equilibrium amount of radium in the mineral has been repeated. I n the first place a solution more than six years old prepared by Boltlwood was found t o be generating radium a t the same rate as i t had done six years prelviously showing that the rate of change of ionium is very small.Four minerals were used (1) Uraninite N. Carolina containing 69.4% uranium very purs and free from alteration products. (2) Cleveite from Saetersdakn Norway containing 10% of rare earths. (3) Broggerite Raade Norway; (4) the same a very pure specimen. The ionium was removed from the solution by precipitating rare1 earths mostly thorium by oxalic acid in (1); by means of hydrofluoric acid in (2); by oxalic acid and thorium in (3); and by rare earths and hydrofluoric acid in (4).Several successive precipitations were employed the final being kept separately as a test solution so that if no radium were generated in it all the’ ionium must have been removed in the previous precipitations. The values for the radio- active constant h(year)-1 and f o r the half-period of radium from these four preparations are shown below H. M. D. I. 11. 111. IV. A X lo4 ..................... 3.71 3-90 4.22 4.14 Half period (years) ... :.. 1865 1777 1642 1774 The last two results are considered the most satisfactory I being a preliminary test and I1 being slightly uncertain on account of an accideat. The mean of I11 and IVY 1660 years agrees excel- lently with Rutherford’s value 1690 years deduced from his experiments with Geiger on the1 counting of the number of a-par- ticles emitted per second per gram of radium.I?. s. The Laws of Radioactive Transformation. A. DEBIERNE (Ann. Physique 1916 [ix] 4 309-322).-The general equation of radioactive change dn2= Alnldt -Ah,n,dt where n1 and n2 are the numbers of atoms of any two successive produds of radioactive constants A and A and the usual resulting integrations are con- sidered from a fresh point of view in which each atom is regarded as spending times t t t . . . in t’he successive forms A B C . . . A so-called ‘‘ equation of destiny” is so obtained which permits the various problelms of radioactive change t o be treated generally in a simple manner. Considerations on the Mechanism of Radioactive Changes and the Constitution of Atoms.A. DEBIERNE ( A m . Physique 1916 [ixl 4 323-345).-It is supposed that in each at<om there exists gtn element of internal disorder. I n the law of F. S.GENERAL AND PHYSICAL CHEMISTRY. ii. 169 a unimolecular chemical change which is formally that of radio- active change the element of disorder is sought in thermal agita- tion. I n radioactive change there is supposed to' be something analogous in the interior of the atom to the decoordinated motion of the molecules of a gas. This idea is discussed generally a t con- siderable length. F. 8. The Distribution of the Active Deposit of Thorium in an Electric Field. (2. H. HENDERSOX (Tmzs. LNovn Scotinn Inrt. Xci. 1914-1915 14 1-16).-A convenient form of apparatus f o r the investigation of the distribution of residual atoms in an electric field is described.Observations made with thorium in dry air show that the1 percentage cathode activity increases with the applied potential the curve expressing the relation having a close resem- blance t o the ionisation saturation curve. With an applied poten- tial of 12,000 volts the cathode activity is 100% within the limits of experimental error and there i& no evidence of the existence of neutral residual atoms. I n ethyl ether vapour on the other hand the whole of the residual thorium atoms are initially uncharged. I n mixtures of ether vapour and air the proportion of charged atoms depends on the composition of the' gas. A similar variation is indicated by the results obtained in experiments with moist air.The Branching Point of t h e Thorium Series. STANISLAW LORIA (Physikal. Zeitsch. 1916 17 6-9).-The result of Barratt and Wood (A 1914 ii 606) puts in doubt the generally accepted scheme of disintegration of the thorium series and the position of the branching point and their explanation and scheme are contrary t o the Soddy-Russel-Fa jans generalisation. The result has been confirmed but the explanation has been shown to be untenable. A well-marked inflection in the curve connecting the volatility of thorium B+C with temperature occurs a t 920° when some 35% has been volatilised. An identical behaviour is shown however by the radium8 curve. Since here the proportion between the two branches is 0.03 99.97 it follows that the explanation advanced as to the inflection a t 35% in the case of thorium namely that the number giving 35% of the a-rays of thorium-C is a distinct sub- stance so that the branching must occur before its formation does not explain the inflection.The cause of this is regarded as of chemical rather than radioactive origin and possibly is due t o the formation of several oxides of thorium-C analogous t o those of the isotope bismuth only stable within certain ranges of temperature. There is a difference in the volatility curves the same for both thorium-G and radium-C according as the active deposit is obtained directly or by electrolysis which mav be due t o formation of alloys with the supporting metal. H. M. D. This view is being further tested. F. S. The Question of Isotopic Elements.11. K. FAJANS (Phzisikal. Zeitsch. 1916 17 l-4).-Thel author now agrees with Hevesy and Paneth in their point of view as regards the potential of aii. 170 ABSTRACTS OF CHEMICAL PAPERS. metal immersed in a solution of its isotope but regarding this as merely a special electrochemizal problem proceeds to criticise the general view of isotopes taken by these’ authors. The chief subject discussed is whether the interpretation of Hevesy and Paneth with regard to their electrochemical experiments is “ new,’’ or differs merely as t o expression from the practice “ a s old as the estab- lishment of the existence of chemically inseparable elements,” of regarding 2 mixture of isotopes as one element which has been repeatedly applied by the author in describing precipitation experi- ments.Only in the case of a metal immersed in a solution of its isotope have He’vvesy and Paneth gone’ further than has been cus- tomary in regarding a mixture’ of elements as one element. Such a system however is thermodynamically indefinite and the line of argument deduced from the application of Nernst’s formula can only be applied with caution. Itl is appropriate t o designate the properties of isotopes not as identical but as equal and there remains the important task of establishing the limits within which this equality holds. F. S. The Question of Isotopic Elements. 111. G. VON HEVESY and F. PANETH (Physikal. Zeitsch. 1916 17 4-6).-With the admission by Fajans of the definiteness of the electrode potential measured by the authors the only dispute as to facts vanishes. Soddy has not raised the question as t o how a system will behave in which the isotopes are not mixed initially and it was only after the authors had .successfully isolated visible quantities of radium-D that such a case could be experimentally studied.This is an impor- t a n t supplementation but no one but Fajans would be under the impression that the authors had ascribed Soddy’s fundamental idea of isotopy to themsellves. Words such as “ our mode of representa- tion” merely arise from the fact that Fajans had attacked not Soddy but the authors. The test now applied-measurement of the electrode potential between pure’ isotopes-enables chemical indi- viduality t o be1 recognised in a much more lucid manner than in the case of the precipitatioas of mixtures of isotopes which are subject to suspicion owing t o adsosption processes and the authors’ results lead t o the conclusion that radium-D and lead are actually “ practically identical in chemical properties.” The authors prefer their term ‘‘ substitutional,” rather than “ identical ” or “ equal ” in the definition of isotopes but to pursue the problem of definition further would be t o cause the controversy to assume the aspect of a mere quibble.F. S. Electrical Resistance of Some Rare Metals ; Thermo-electric Power and Rectifying Action of Germanium. CARL RENEDICKS ( I n fern. Zeitsch. illetallographie 1916 7 225-238).-A specimen of massive cerium free from pores is found to have a specific resistance of 78.Praseodymium and neodymium give the values 88 and 79 respectively. The specific resistance1 of ruthenium is probably about 9 the actual value1 found beling higher on accountGENERAL AND PHYSICAL CHEMISTRY. ii. 171 of porosity. Uranium has a lower resistance than bismuth but the metal is not readily obtained free from carbide. The specific resistance of germanium lies between the values for silicon and tin. Measurements of the tlierrnoelectric power of gerinanium give as the value of the coefficient a against copper - 380 rt 40 microvolts per degree. Germanium has thus next to silicon the highest known thermoelectric force against copper. The fact t h a t silicon occasionally gives positive values is attributed to allotropy as grey tin is found t o be1 positive whilst ordinary t i n is negative.Germanium like silicon exhibits a marked rectifying effect. The thermoelectric current passes much more readily in one direction than in the other especially when the contact surface is fresh and the pressure light. This happens even when the point and plate in contact are of the same metal. The direction of the effect makes a purely thermoelectric explanation inapplicablel. C. H. D. Contact Electricity of Solid Dielectrics against Conducting and Non-conducting Liquids. ALFRED COEHN and JOSEF FRANKEN (A72n. Physik 1915 [iv] 48 1005-1033).-A method is described by means of which it. has been found possible to measure electrostatically the1 charge which solid dielectrics acquire when brought- into contact with aqueous solutions The contact electrical effect increases with time but' reproducible end-values are readily obtained .Solid paraffin was employed as dielectric substance and the influ- ence of electrolytes and non-electrolvtes on the contact effect was examined. The influence of electrolytes on the charge acquired by the dielectric is similar to that' which has been previously found in experiments with gases bubbled through the aqueous solutions but in the case of paraffin the electrolyte concentrations required f o r a given reduction in the magnitude of the contact effect are1 much greater than in the case of eases. This parallelism shows itself in the much greater influence of the hvdroqen ion as compared with other ions. I n all the experiments with acids i t was found possible to bring about a reversal of siqn of the contact.effect 7nv increasinq the concentration of the solutions. The difference in behaviour between acids and neutral salts is also shown in the temperature-coefficient' of the contact effect in t h a t temperature has little influence in the case of neutral salts and alkalis whereas the influence of acids increases rapidly with rise of temperature. Non-electrolytes in dilute solution have no appreciable influence cn the contact effect; a resnlt which was also obtained in experiments with gases. H. M. D. The Electrical Resistance of Acetic Acid in the Solid and Liquid Phases. J. H. L. JOHNSTONE (Tmizs. ATOWL Scoticcu Inst. Sci. 1912-1913 13 191-208).-The influence of temperature on the electrical resistance of acetic acid has been examined over the range - 8 O O to 27O and the influence of small quantities of water on the resistance determined.The acetic acid was pre'pared by fractional freezing of a sampleii. 172 ABSTRACTS OF CHEMICAL PAPERS. which contained 99.55% of acetic acid but it is not claimed t h a t water was completely eliminated by this treatment. The presence of water is indicated by thel fact t h a t a sharp change in the resist- ance was observed a t the eutectic temperature (- 26.55O) as well as a t the melting point of the acid. For a particular sample! of acid the resistance changed suddenly from 1O1O below to 8.8 x 108 above the euhctic temperature and from 1-27 x l o 9 below to 3-52 x lo7 above the) m. p. The variation of thel resistance in the interval between the eutectic temperature and the m.p is anomalous b u t can be explained in terms of the changing concentration of the liquid phase which is present. The conductivity ob the acid in this region is due almost entirely to the prenence of traces of water. H. M. D. Galvanic ( ( Exaltation” of Metals by Alcohol. CII. M. VAN DEVENTER (Ghem. TVeelcblad 1916 13 173-183).-A further con- tribution to the author’s theory of the leaking insulator. A. J. W. New Thermoelectric Method for the Study of Allotropy of Iron or Other Meta-1s. C. BENEDICKS (Cowzpt. re?zd. 1916 162 297-299. Compare) J . Iron Steel Inst. 1914 1 434).-The metal was examined in the form of a fine wire which was drawn a t a constant. velocity of 1.6 mm. per second through a small electric furnace maintained a t a constant temperature measured by a Le Chatelier couple.Ileasurements were made of the E.M.F. developed a t different. temperatures. The curve for the sample of iron examined which was remarkably pure showed a marked dis- continuity corresponding with the point A b u t for the point A no discontinuity was found. The curves obtained for this effect resemble the dilatation curves. W. G. Magneto-chemical Effect. A. N. SCHTSCHUKAREV ( J . Rws. Phys. Chem. SOG. 1915 47 1644-1668).-The author assumes t h a t tho process of chemical combination or in general of the mutual action between two impinging molecules is equivalent to an electric discharge so t h a t the medium in which such chemical process takes place will represent a collection of such discharges proceeding in all possible directions.In a magnetic field there should be an increase in the number of discharges perpendicular to the lines of the field and therefore a temporary difference’ of potential in the same direction. I n a vessel situate in a magnetic field were arranged parallel t o the lines of the field two similar pieces of platinum foil perfectly identical as regards electromotive force’. When these plates were connected through a galvanometer and the vessel was charged with ferric chloride solution no deflection was observed but addition of potassium iodide to the liquid caused an immediate deflection. This phenomenon of chemical polarisation accompanies many other and possibly all reactions. As a rule the smaller of the platinum plates becomes negatively and the larger positively electrified.With one and the same pair of platinum plates hhe deflection gradually diminishes owing to their exhaus- tion but the latter effect may be reduced by occasional immersionGENERAL AND PHYSICAL CHEMISTRY. ii. 173 of the plates in concentrated hydrochloric acid. Thus the initial magnitude of the current of chemical polarisation varies but in every case the current exhibits a very gradual diminution which probably extends over the whole time occupied by the reaction. The existence of the current of magnetic polarisation referred to1 above has been actually observed (1) with reactions proceeding freely and (2) with reactions taking place a t electrodes during electrolysis. The results obtained appear t o indicate a connection between this magnetic effect and diminution in valency.The magnetic effect is indeed o'bserved only with reactions in which there takes place a lowering of thO valency of one of the reacting ions such as occurs in the reactions between ferric chloride and potassium iodide chromium trioxide and hydriodic acid and vana- dium tetrachloride and potassium iodide. If hcwever the libera- tion of iodine takes place notl on account of the depression of the' valency of one of the reacting ions but' on account of the decom- position of a complex ion o r a non-ionised compl.~x or if the diminished valency of one of the reacting ions is compensated by increased valency of another as in the reaction 2FeCl,+ SnCl no magnetic effect is observed. Examinat'ion of a number of cases in which the magnetic effect accompanies electrodic changes leads t o the following coiiclusions Complex ions which on separation react only with water such a s OH NO SO CH,*CO CN S,O ClO give no magnetic effect but when ions f o r instance SO separating from an -ous salt unite immediately with the dissolved salt forming an -ic salt a marked magnetic effect is observed.Zn" and Fe" give no magnetic effect but Fe"' gives a very strong effect which changes its sign. With rare exceptions the1 sign of the magnetic effect is constant f o r a given reaction o r for the separation of a given ion. It appears strange that Ca" gives no effect whereas with Mg" the effect is very marked. Particularly int'ense magnetic effects are given by reactions accompanied by separation of ions of the halogens ; with bromine and iodine an almost constant negative1 effect is observed with a normal field and this is also the case with the separation of chlorine ions from a whole series of salts although with another series of chlorides a marked positive effect is observed.No effect is produced in cases where the liberated chlorine undergoes immediate combination; this is the case f o r instance with the reaction SnCl + C1,= SnCl or when a small proportiop of ally1 alcohol is added to a solution of barium chloride. T. H. P. Thermal Expansion and Compressibility of Liquids at Low Temperaturea. W. SEITZ H. ALTERTHUM and G. LECHNER (Ann Physik 1916 [iv] 49 85-92).-1n the expectation that the pro- perties of liquids would show a tendency t o approximate a t low temperatures to those characteristic of solids measurements have been made of the' change of the specific volume of a number of liquids o'f low freezing point.For isopentane and ethyl ether the coefficient of thermal expansion diminishes with falling temperature down t o -120° the lowest temperature a t which observations were 8*ii. 1‘14 ABSTRACTS OF CHEMICAL PAPERS. made. For methyl alcohol and ethyl alcohol the coefficient decreases a t first attains a minimum a t about -40° and a t lower tempera- tures increases. Carbon disulphide was found t o behave similarly the minimum in this case corresponding with a temperat’ure between -70° and -8OO. H. M. D. Thermal Expansion and Compressibility of Liquids at Low Temperatures. W.SEITZ and Q. LECHNER ( A ~ L ~ L . Yhysik. 1916 [iv] 49 93-115. Compare preceding abstract).-The compressi- bility of isopentane ethyl ether methyl alcohol ethyl alcohol and carbon disu!phide has been determined a t temperatures between Oo and -110” and at pressures up to 1000 atmospheres. As the pressure increases the tendency of the coefficient of expansion t o increase with falling temperature becomes much more marked than a t atmospheric pressure. H. M. D. [Determination of the Velocity of Sound in Gases.] C. DIETERICI (Ann. Physik 1915 [iv] 48 1122-1124).-A reply t o Schweikert’s criticism (this vol. ii 79) of the results obtained in previous measurements of the velocity of sound in gases. The results in question are shown t o be in good agreement with Schwei- kert’s own data.H. M. I>. A Criticism of van der Waals’s Equation and Some New Equations Derived Therefrom. JAMES KAM (Phil. Mag. 191 6 [vij 31 22-36).-From a consideration of the influence of a and 6 on the pressure-volume relations the author arrives a t the formula Y+a/(V+ b)2=R2’/ V as an alternative to the van der Waals’s equation. According to this formula an actual gas of density 1 / (I’ + 0) exerts the same pressure1 as an ideal gas of density 1/ V . The analysis of the formula leads to the conclusion that the density in the critical condition is for all substances two-thirds of the density of an ideal gas which a t the same volume exerts a pressure equal to the critical plessure a t the critical temperature. If the internal pressure is denoted by Pi so that the total pressure ir = P + Pi then i t is shown that the critical values of P and 1’ are equal or Y =PI$ and further that Y = 12T,/d Vc and 7rc = 11!f’,/ V according t o which the ideal gas law is applicable t o the critical condition.Experimental data in support of the validity of this relation are cited. The equality of P and Pi is supposed t o be connected with the disappearance of the surface tension and of the latent heat of vaporisation a t the critical point. When this equality obtains the cohesive f o r e s are exactly counterbalanced by the thermic pressure and the transference of a molecule from the interior of the liquid t o the space above it requires no expendit,ure o,f energy. H. M. D. The Variation of Surface Tension with Temperature.ALLAN FERGUSON (Phil. Mag. 1916 [vi] 31 37-47).-1t is shown that the variation of the surface tension of liquids with temperatureGENERAL AND PHYSICAL CHEMISTRY. ii. 175 can be represented by the! equation T=T,(l -bB)n in which T is the surface tension a t t'emperature 6 To the tension a t Oo and b and ? L are constants characteristic of the substance. Since T = O a t the critical temperature 6 it follows that 8 = 1 / h. The critical temperatures calculated in this way are found t o be in close agree- ment with the) observed values f o r some fourteen non-associated liquids which have been examined. The value of h varies from 0.002793 f o r clilorobenzene to 0.005155 f o r ethyl ether but 'IL varies very little from the mean value of 1.210. This mean value may be employed generally without seriously detracting from the accuracy of the formula.I n the group of esters R-CO,R the value of b diminishes with increase of R o r R and is greater f o r an iso-compound than for the corresponding normal ester. Moreover R and R cannot be intar- changed without alteration in 6 and this has the greater value when the more complex radicle is associated with the carbonyl group. I n esters of the series R-CO,Me the value of b is apparently con- nected with the boiling point 6 of the ester as expressed by the equation b = c / ( d - ? ~ ) 6 - 273c in which c and d are constants and / I is the number of carbon at'oms in R. The empirical relations connecting the temperature-coefficient of the surface tension of a liquid with its critical temperature which have been pointed o a t by Walden (A.1909 ii 1221 are discussed in relation to the author's formula. H. M. D. Thermodynamics and Fractional Distillation. E. CHENARD (Bull. Assoc. chim. Suer. Dist. 1916 33 47-55).-A discussion of the influence of various factors on the separation of the components of binary liquid mixtures by fractional distillation with special reference t o mixtures of ethyl alcohol and water. It is pointed out t h a t the rate of flow of the vapour and the difference in density of the components exert an influence on the fractionation which is of practical importance and may lead t o results which are quite different from those anticipated from a consideration of the thermo- dynamic equilibrium between thel liquid and vapour phases.For the construction of improved forms of distillation apparatus i t is shown that the design must be such as t o permit of free passage of the vapour and rapid removal of the condensed liquid. The frac- tionating column should be disposed horizontally or nearly so and the number of fractionating elements should be considerable. Actual elxperiments with apparatus constructed on these lines have shown t h a t much better results can be obtained than by the use of apparatus of the' usual type. H. M. D. Heats of Formation of Additive Organic Compound@. IV. Picrates. B. L. VANZETTI rmd V. GAzz.4~1~ ( 8 t t i R. Accnd. Lime; 1915 [v] 24 ii 527-532; Gnzzettn 1916 46 i 145-151. Com- pare A. 1913 ii 296).-Them heats of formation of the picrates of certain bases of the pyridine group now given indicate clearly the differences existing between compounds of picric acid with tertiary amines s ~ c h as pyridine and quinoline and those of the same acid 8"-2ii. 1'76 ABSTRACTS OF CHEMICAL PAPERS.with the corresponding hydrogenated secondary basw in which the nitrogen-containing group has lost completely the aromatic char- acter. The following are the mean molecular heats obtained in large caloriels pyridine picrate (m. p. 164O) 13-84 ; piperidine picrate {m. p. 151-152O) 20.56; quinoline picrate (m. p. 204O) 13.2; tetrahydroquinoline picrate (m. p. 141*5O) 9.9. Physico-chemical Force of Attraction. 11. N. A. KOLOSOVSKI ( J . Buss. Phys. Chem. Soc. 1915 47 2035-2037).-Reply to Gur- vitscli (A. 1915 ii 747). HENRY LE CHATELIER (Cornpi.rend. 1916 162 245-246).-A reply to Colson (compare this vol. ii 129). W. G. T. H. P. T. H. P. The Law of Solubility. Ultramicroscopy of Crystallisation Phenomena. I. P. P. VON WEIMARN ( J . Buss. Phys. Chem. Soc. 1915 47 2140-2163).- After discussing the capacity of highly disperse particles for vec- torial growth etc. experiments are described on the formation and disaggregation with the disperse particles of skeletons of sodium chloride in silicic acid gels. From the results obtained under different experiineutal conditions and the microscopic and ultra- microscopic appearance of the crystalline systems obtained a number of general conclusions are drawn. Crystalline skeletons and dendritic growths resulting from the prevention of regular crystal- line development by marked initial supersaturation or obtained in a disperse medium may be regaxded as coalitions of disperse par- ticlee which are to some extent irregularly orientated t o one' another. These1 crystalline systems in consequence of the great delvelopment of their external and internal surfaces the latter the result of rupture of the crystalline' homogeneity exhibit instability and break down into disperse particles od a different degree of dispersion ; this disaggregation is accelerated by small oscillations of the temperature.The disperse systems thus formed undergo gradual diminution of their dispersion and the disperse particles then tend t o consume) one another and thus yield crystals with com- plete outlines. The capacity for vectorial growth shown by large crystals is inherent in particles of all degrees of dispersion includ- ing those a t the extrelme limit of ultramicroscopic visioln.Precipi- tates formed by the disaggregation of dendritic systems and consist- ing of particles which do not reveal their crystalline form when examined ultramicroscopically are still undoubtedly crystalline. When dissolved these crystalline formations characterised by marked development of their external and internal surfaces o r by rupture of the crystalline homogeneity undergo disruption into disperse particles this procws preceding solution. T. H. P. Materials for Experiment ~1 Dispersoidology. I. Prepara- tion of any Substance in any Degree of Dispersion. P. P. VON WEIMARN ( J . Russ. Phys. Chem. SOC.1915 47 2133-2139).-The author d i s c u m the colloidal condition a8 a general property ofGENERAL AND PHYSICAL CHEMISTRY. ii. 177 matter (compare A. 1910 ii 940) and describes again the elxperi- ments with sodium chloride (A 1913 ii 31). Further results are given in succeeding papers. T. H. P. Materials for Experimental Dispersoidology. 11. Gels and the Process of Gelatinisation. P. P. YON WEIMXRN ( J . Rziss. L'hys. Chew. SOC. 1915 47 2163-2176).-m7hen a gel is formed by the more or less uniform gelatinisation of a liquid throughout its entire mass the. author terms i t reticulated and whe'n the gela- tinisation takes place at' definite surfaces within the liquid coarsely- cellular. A gel of the1 latter type is obtained f o r instance# when saturated aqueous barium thiocyanate is added t o saturated man- ganous sulpliate solution every drop or stream of the added liquid becoming coated with a transparent membrane of gel.Systems analogous t o coarsely-cellular gels are obtainable with ice by pouring drops of water into alcohol cooled in a mixture of carbon dioxide and ether o r in liquid air; it is difficult t o fix the gelatinous form of the ice membranes for any length of time but the structure of these1 membranes may be examined microscopically and ultramicro- scopically and is found to be identical with that of gelatinous mem- branes of barium sulphate. The formation of membranes in this case is due t o the fact that a t such low temperatures the water is able to diffuse scarcely a t all into1 the alcohol and is almost instan- taneously solidified in a highly disperse state.Reticulated gels of ice are obtainable particularly easily near the' eutectic point of the water-alcohol syskem. Membranes of coarsely-cellular gels are closely comparable with sections of correspoading thickness of reticu- lated gels cut by means of a microtome. Many observations confirm the view that the process of gelatinisation and the formation of distinctly niicrocrystalline precipitates are essentially identical and differ only in degree. The swelling spontaneous dispersoid solution and gelatinisation of sodium oleate in toluene xylene and benzene have been investi- gated the observations made being in complete agreement with those' of the same processes with barium sulphate gelatin alumin- ium hydroxide etc.confirming the conclusion that a gel is a sponge composed of highly disperse crystalline granules soaked in dispersive medium. T. H. P. Materials for Experimental Dispersoidology. IV. Disperre Systems of Cupric Chloride in Benzene P. I?. VON WEIRIARN and I. B. KAGAN ( J . Rms. Phy:. Ghem. SOC. 1915 47 2215-2251).-The conditions of formation and certain properties of the following disperse systems have been investigated With the liquid disperse phase zH,O + yHC1; with the solid disperse phase CuCl ; with the solid disperse phase aCnC1 + yCuCl2,2H2O ; with the solid disperse phase CuC1,,2H,O this being the first example found of a typical crystallo-hydrate in the disperse state; with t,he solid disperse phase CuC1,.2H,O and the complex liquid disperse phase (rH,O + yMCl + zCuC1,) and (xR,O + yHC1) ; with the solid disperse phase (xCuC12 + y copper oleate) ; with the solid disperseii.178 ABS'l'ICACTS OF CHEMICAL PAPERS. phase (xCuC12,2H,0 + y copper oleate). Confirmation is obtained of the theory according to which in a medium either almost com- pletely passive o r very active the stability of dispersoid solutions is small. The colour of dispersoid solutions is nearly identical with that of true solutions and that of coarse disperse phases but is less intense than thel latter and with a sufficiently high degree of dispersion is turbid; i t is hence possible t o judge by the colour of the course of chemical and physical transformations of disperse phases. T. H. P. Nature of the Elastic and Plastic Conditions of Matter.I. I. OSTROMISSLENSKI (2. Russ. Phys. Chern. SOC. 1915 47 1995-2014) .-The two new constants for colloids namely the elasticity point and the fatal temperature (this vol. i 54 55) permit of the characterisation of two new states of aggregation of matter (1) the plastic state in which a substance exists in the interval of temperature between the elasticity and fatal points and (2) the elastic statel corresponding with the interval between the elasticity point and the melting point. The supposition that these states of aggregation are properties of all colloids is confirmed experimentally and since the colloidal condition is common t o all forms of matter the elastic and plastic states are general forms of aggregation. The normal course of the changes in the condition of aggrega2n is expressed by the scheme solid plastic t elastic +- liquid S gats.. Many substances follow exactly this scheme1 of transformations bef ore1 reaching the temperatures a t which their molecules undergo decomposition but others melt with previous conversion into only one of these intermediate states or into neither of them.Such substances correspond with the following in- complete schemes solid = plastic = liquid solid = elastic f liquid o r solid liquid and cannot be1 obtained in the plastic or elastic condition by simple change of t'emperature. The methods availablel f o r obtaining matter in the' plastic o r the elastic state are1 as follows (1) By changel of temperature. Various individual compounds and mixtures such as caoutchouc and some of its homologues polymeric acrylic esters factis gels of gelatin etc.exist in the1 elastic condition at the ordinary temperature and others such as guttapercha camphor copper oleate paraffin wax ozokerite etc. in the plastic state. The transformation from the elastic t o the plastic state is always realisable since i t accompanies lowering of temperature but that from the plastic to the elastic condition by rise of temperature does not always occur since fusion o r decomposition may take place before the elasticity point is reached. I n such cases one of the two following methods mustl be employed. (2) By introduction of a foreign liquid. This liquid must be able to act as a medium in which the given colloid elither dissolves o r swells the process of solution being intimately con- nected with the plastic condition of matter and that of swelling with the elastic state.Solids which neither dissolve nor swell in liquid media cannot be obtained in elither the elastic or plastic state by this method. For example caouprene chloride readily swells inGENERAL AND PHYSICAL CHEMISTRY. ii. 119 carbon disulphide a t the' ordinary temperature giving an elastic substance possessing all the properties of natural caoutchouc but exhibiting instability owing t o the ease with which the carbon disulphide is lost. Again the1 compound of isoprene with sulphur dioxide (S02)nln(C5H8)S is a typical solid with a fatal temperature above the ordinary temperature; when heated i t is converted first into the plastic condition and then approximately into the elastic form but a t this point decomposition begins.If however this compound is precipitated by alcohol from its solution in hot aniline i t yields the complex n(C,H,),,(SO,) + pNH,Ph + cyEt*OH which a6 the ordinary temperature exhibits qualitatively and quantita- tively the association of elastic properties of ordinary caoutchouc its elasticity point being below Oo. (3) By supercooling liquid masses. By rapid and intense1 supercooling for instance t o - looo a liquid or fused solid may be obtained in an elastic o r plastic modi- fication a t a temperature corresponding normally with the solid form alone. This method is generally known and is used princi- pally for obtaining substances in amorphous solid vitreous modi- fications; the plastic and elastic states represent intermediate stages.Thus by rapid and profound cooline water is obtained first in the elastic then in the plastic and finally in the vitreous modification and similar behaviour is shown by almost all liquids. Substances in the plastic o r elastic condition are not to be regarded as mere mechanical mixtures of a solid and a liquid although their peculiar properties are due t o mutual action of a solid and liquid. When a liquid o r fused substance is rapidly and intensely cooled the low temperatiire1 causes reduction in the velocity with which its solid phase appears so that part of the' substance retains its liquid consistency ; into this portion the nascent solidified part swells the elastic o r plastic modification being thus obtained a t temperatures correspondinq normally with the solid phase.The following principles are deduced (1) The trans- formation of a solid into a liquid usually commences considerably below its melting point. (2) Every given temperature of a sub- stance corresponds with a definite stable equilibrium between its solid and liquid phases. With this equilibrium there corresponds with any substance a definite association of elastic o r plastic pro- perties ; in particular f o r a certain temperature interval the propor- tion of one of the phases may be zero. A colloidal solution of a solid substance! containin? the latter partly in the solid state should when heated behave in a definite manner independently of the nature of the solvent. Since also with everv temperature there corresponds a definite equilibrium between the solid and liquid phases the solid will be Converted successively into the plastic elastic and liquid states indenendentlv of the condition of xqgregation of the sin-roundinq medium; this Drocess should therefore take place both in a gaseous medium and in colloidal solution.It has been already foiind that the elasticitv point amd the fatal temnerature may be determined by means of either the substance in t.he free state or its colloidal solution. The melting points of colloidal substances are also measurable in anii. 180 ABSTRACTS OF CHEMICAL PAPERS. analogous manner t h a t is by the change in direction of the curves expressing the dependence of physical properties on the temperature (this vol.i 55). Tlie temperature thus determined would corre spond with the disappearance of the solid phase in the given colloidal solution and in reality it may be accompanied by either (1) separa,tion of the solution into two layers or (2) its conversion into a molecular disperse solution o r (3) the formation of a colloidal solution of a liquid in the same solvent. T. H. P. Non- Uni- and Bi-variant Equilibria. IV. F. A. H. SCIIREINE- MAKERS (Proc. K . Akad. TVeteilzsch. Amsterdam 1916 18 1018-1025. Compare this vol. ii 19).-A further consideration of types of equilibrium in heterogeneous systems with special refer- ence t o the general case of a syste'm of ?z components. The n + 2 equations which are required to define the univariant equilibria are not independent of one another.TWO of these equations only are required for the determination of the remainder. H. M D. Non- Uni- and Bi-variant Equilibria. V. F. A. H. SCIIREINE- MAKERS (I'roc. K . Akad. TVetensch. A msterdam 1916 18 6026-1037. Compare preceding abstract).-It is shown t h a t the different types of pressure-temperature diagram corresponding with any system of TL components may be deduced without a knowledge of tlik concentration diagram or the composition of the phases. H. M. D. Determinations in the System Lead Sulphate Sulphuric Acid and Water. A. D. DONK (Chenz. TVeekbZad. 1916 13 92-97).-An application of Schreinemakers's graphic method t o mixtures of lead sulphat,e and dilute sulphuric acid. A. J. W. Swelling of Hides in Presence of Hydrogen Ions.G. POVARNIN ( J . Buss. Phys. Ghem. Soc. 1915 47 2064-2073).-The work of other investigators notably Paessler and Appelius (Gerber- Zeit. 1902 45) in this direction is discussed. Tlie velocity with which hide-material swells in pure water is expressed by the equa- tion log [ M / ( M - Q)] = 0 4 3 4 3 4 t (l) and the swelling in presence of acid depends on the following considerations ( a ) The velocity of swelling is directly proportional t o the velocity of the acid ions in penetrating the membrane (hide) and this velocity is propor- tional to d? as was shown by Morsel and Pierce (A. 1904 ii 14) for gelatin; in absevlce of acid the swelling is proportional t o the velocity with which the ions of water pass through the membrane t h a t is t o t . ( b ) The' rate of swelling is inversely proportional t o the quantity of the1 hide-material dissolving a t the surface of contact of hide and liquid and since' this swelling is directly pro- portional to the time the rate of swelling is invers2l-y proportional to t . ( c ) The maximum swelling Jf for a given acid is equal to some constant maximum M multiplied by Lo which is a ccefficient varying with the nature of the acid; it may be assumed that k1 of equation (1) is in general equal t o some magnitude k/2.(d) TheGENERAL AND PHYSICAL CHEMISTRY. ii. 181 rate of swelling is proportional t o the index of basicity n of the acid if the latter is completely dissociated and in general it should be proportional t o the number of free hydrogen ions in the solution of the given acid.On these1 theoretical foundations the equation log [ X / (ill - Q)] = 0.4343nJc d j is constructed. It is found that Paessler and Appelius's results for five acids a t concentrations of 0*1-0*6% are' expressed satisfactorily by this equation 0.4343k having the value 0.274 whilst n is equal t o 1 or with strongly dissociated dibasic acids 2 ; M is a variable magnitude depending on the character and concentration of the acid. T. H. P. Interaction of Benzoyl Chloride a.nd m-Xglene in Presence of Haloids of Various Metals of the Second Group. B. N. MENSCHUTKIN ( J . Rprss. Phys. Clzem. SOC. 1915 47 1853-1884).- By absorption of the hydrogen chloride liberated in potassium hydroxide solution the author has measured the velocity of the reaction Pli-COCl+ m-C?H4Me = C,H3Me2Bz + HC1 a t 1 5 8 O in pres- ence of varying proportions of hzloids of metals of group 11.The following considerations have a bearing on the results obtained. When antimony trichloridel or tribromide is present (compare A. 1914 i 188 673) this forms a molecular compound with the aromatic hydrocarbon readily soluble in the latter so t'hat the re'action with the chloroanhydride proceeds in a perfectly homo- geneous medimi. The halogen salts now investigated with the possible exception of tQhose of zinc and mercury are quite insoluble in either benzoyl chloride o r sn-xylene o r in mixtures of the two and as far as can bel seen enter into reaction with neither of the relacting compounds; in these cases then the media are hetero- geneous. Further although most of the salts are kept thoroughly mixed with the liquid by a current of air passing through the latter this is not the case with such heavy salts as mercuric chloride and bromide; in these instances then the values of the velocity constiants obtained are lower than the true values.The values of the velocity constant obtained when approximately 0.5 o r 1.0 mol. of the haloid is taken per 1 mol. of each of the reacting compounds are as follows the first. figure in each case giving the absolute value followed by the relative value in brackets lCaCl X 0.0515 (1) ; 0*94CaBr 2Kz 0.0514 (I) ; 0*98SrC1 K2 0.0563 (5) ; 0*5SrBr X 2 0.0423 (16) ; 0*48BaC1 X 0.0578 (6) ; O*5BaBr K 0.0413 (9) ; 0*57CdC1 KO 0*0462 (44) ; O*5lCdBr X 0.0417 (12) ; 0*49MgCI K 0.0415 (11) ; h*51HgCI2 KO 0.0334 (243) ; l-OSHgBr K 0*0245 (3214) ; 0*53ZnCI R 0.0110 (7857) ; 0*52ZnBr2 R 0.0121 (8643).I n presence of the zinc and mercuric salts the velocity constant increases only t o a maximum value given above these maxima being obtained with 0*38ZnC12 0*26ZnBr 0~18HgCI and 0*26HgBr respectively. The values of K2 corresponding with the presence of 1.03 0.53 0.5 0.38 0.27 0.105 and O*O13ZnCl are 0.0125 0*0110 0.0106 0.0112 0.0071 0.0047 and 0.00021. The variation of IT with the concentration of the halogen salt is here quite different from t#hat found when varying proportions of antimony trichloride are present since K is then almost exactlyii. 182 ABSTRACTS OF CHEMICAL PAPERS. proportional to’ the square of the concentration of the trichloride (Zoc.cit.). The use of zinc chloride and bromide in organic syn- theses is now undelr investigation. T. H. P. Velocity of Decomposition of Electrolytes in the Light in Relation to their Degree of Electrolytic Dissociation. C. BONGIOVANNI (Gnzzetta 1916 46 i 127-135).-The fact that the instability of perchloric and nitric acids is not shown by aqueous solutions of these acids o r by the corresponding salts Ostwald (“ Principles of Inorganic Chemistry ”) regards as due to the occur- rence in the latter cases of the stable perchloric or nitric ion which is not present in the anhydrous acids. The author considers this conclusion unfounded and points out that water often has a positive or negative1 catalysing influence on the decomposition of compounds dissolved in it and further that the latter often combine with the water yielding compounds differing in stability from the anhydrous compounds.A number of cases are quoted in support of this view. The results of the author’s own experiments are as follows. -* I The velocity of decomposition of oxalic acid in dilute aqueous solution by the action of sunlight is considerably diminished by the presence of sulphuric acid and the depression of the freezing point of the solution containing the two acids is far less than the sum of the depressions observed with solutions of the two acids taken separately. According to Ostwald’s hypothesis similar behaviour should be shown by solutions of oxalic acid and potassium oxalate. I n this case however the presence! of potassium oxalate causes marked retrogression of the degree of dissociation of oxalic acid but this is not accompanied by any variation in the stability of the acid towards sunlight.Thew results appear t o exclude1 the assumption that the molecule1 of oxalic acid exhibits a stability towards light different from that of its ion. T. H. P. A Kinetic View of Catalysis. H. J. PRINS (Chenz. Weekblnd. 1916 13 127-131).-A theoretical paper. A. J. W. The Influence of Different Surfaces on the Decomposition of Methane. WILFRID ERNEST SLATER (T. 1916 109. 160-164) ,-The influence of different surf aces on the thermal decomposition of methane has been investigated in comparative experiments a t 910° in which the gas was exposed t o equal surfaces of different substances f o r equal intervals of time and the decom- position measured by the amount of hydrogen in the1 residual gas.The various substances examined were used in the form of fine powders which had been sifted through fine gauze and were disposed in two porcelain boats placed end to end along the middle portion of the electrically heated porcelain tube. The results obtained with silica alumina magnesium oxide calcium oxide barium oxide wood charcoal graphite csrborundum iron and copper show that the rate of decomposition of methaneGENERAL AND PHYSICAL CHEMISTRY. ii. 183 depends not only on the e’xtent of the hot surface which is exposed to the gas but; also on the nature of the substance concerned. H. M. D. Velocity of Hydrogenation of Fumaric Acid with Colloidal Palladium as Catalyst.A. KOREVAAR (C7zcm. Weekblad. 191 6 $13 98-107).-The author points out that the velocity of hydro- genation of fumaric acid with colloidal palladium as catalyst is dependent on the rapidity of revolution of the stirring apparatus and describes a form of stirring gear with which the velocity of revolution can be measured and controlled. A. J. W. The Metastability of the Elements and Chemical Com- pounds in Consequence of Enantiotropy or Monotropy and its Bearing on Chemistry Physics and Technics. I. ERNST COHEN (Proc. I!. ,4 Icad. Wetenseh. Am.sterdnm 1916 18 961-965). -Observations on the enantiotropic yellow and red modifications of thallous picrate are described which show that the dry substances may be heateld o r cooled far above o r below the transition tempera- ture without change.I f traces of a solvent such as water ethyl alcohol methyl alcohol o r ether are added the retardation pheno- mena are not observed. It is suggested that arrested transformation may be of frequent occurrence amongst the non-metals and chemical compounds and that’ this property is not confined to the metals. Integral Atomic Weights. I. and 11. FRANK WILLIAM DODD (Trans. iVoua Scotia Tnst. Sci. 1912-1913 13 216-221 223-227). -1. It is suggested that the properties of the elements are func- tions of integral atomic numbers which are greater than the accept’ed atomic weights t o an extent depending on the density of the element in accordance with the relation I . N-A . W = 1.8D where I . N is the integral atomic number A . W the atomic weight and l3 the density. Such properties are the specific heat and the m.p. 11. To meet the1 criticism that the relation between integral atomic number atomic weight and aensity will depend on the temperature at which the density is measured i t is suggested that each element has a natural density any departure from which is more or less accidental. LEONARD DOBBIN (Chem. News 1916 113 85) .-Avogadro’s hypothesis (1811) which was not generally accepted until some years after the publication of Cannizzaro’s ‘‘ Sunto di un corso di filosofia chimica ” in 1858 was clearly stated by Prout in his “Chemistry Meteor- ology and the1 Function of Digestion considered with reference t o Natural Theology,” which was published in 1834. Prout’s views which were adopted without knowledge of the essays of Avogadro Arnp&re and Dumas were criticised by W. C. Henry (Phi7. Mng. 1834 [iiil 5 33) but were not otherwise noticed until 1904 (see A. N. Meldrum (( Avogadro and Dalton ”) H. M. D. H. M. D. Prout in Connexion with Avogadro’s Hypothesis. N. H. J. M.ii. 184 ABSTRACTS OF CHEMICAL PAPERS Valency of the Elements. 111. G. POVARNIN (J. Rz~ss. Phys. Chem. SOC. 1915 47 1737-1769. Compare A. 1915 ii 761).- The author develops further his theory of polar affinities and dis- cusses the views of Tschitschibabin who states (“ Investigations on Teirvalentl Carbon”) ‘‘ Valency is a number denoting how many atoms are in immediate union with the given atom.” “Valency and also the degrees of saturation that is tlie store of chemical energy of an individual atom is a function of the atoms and groups entering into Combination with such atom.” “ The strength of the separate valency of each atom is not a constant magnitude but varies . . . in dependence on the remaining radicles united with the atom.” The insufficiency of these views is maintained parti- cularly with reference tot various points connected with derivatives of “ tervalent ’) carbon and i t is claimed that the author’s hypo- thesis gives more reasonable explanations of such points. T. H. P. Stesm Apparatus with Constant Water Supply. J. TRAMBICS (Chem. Zed. 1916 40 128-129).-A simple apparatus consisting of a boiler of thin metal provided with a constant level tube similar t o that used for water-baths except that’ i t is attached to’ the water- tap and has two tubes cdnnecting it with the boiler one! below and the other above tlie water level; the outlet is connected with a U-tubs which gives a pressure in the apparatus corresponding with the column of water. A mercury regulator is employed for tlie gas supply. A sketch of the apparatus is given. N. H. J. M.