69 General and Physical Chemistry. Enhanced Lines of Titanium Iron and Chromium in the Fraunhoferic Spectrum. Sir J. NORMAN LOCKYER and F. E. BAXANDALL (Proc. Roy. Xoc. 1904 74 255-267).-The contents of this paper are summarised as follows. The enhanced lines of titanium and iron are practically all represented in the Fraunhofer spectrum but in some cases the corresponding solar lines are compound and only partly due to one or other of these metals. The corresponding solar lines are generally speaking comparatively weak ones The majority of the chromium enhanced lines occur in the solar spectrum although some appear to be missing. Some of the Fraunhofer lines correspond with metallic lines special to the spark spectrum and lacking in the arc and probably for this reason they were left unorigined by Rowland.J. C. P. Spectrum Regularities and the Atomic Weight of Radium. GEORGE RUDORF (Zeit. pIt,ysiknZ. Chew,. 1904 50 100-1 10. Compare Runge and Precht Abstr. 1903 ii 346 ; Watts Abstr. 1904 ii 720). -The author finds that frequency differences (d) and atomic weights ( A ) are not exactly related according to the equation d/A” = const. ; hence the formula logA = CL + b log d is not exactly valid and any con- siderable extrapolation based on the straight line formula may lead to error. It is shown that certain graphical relationships exist between the values of A and 100d/A2 vhich permit of the calculation of d for other elements. The value 225 should for the present be taken as the atomic weight of radium. J . C. P. Racemism. GIUSEPPE BRUNI ( A t t i I?.Accad. Limei 1904 [v] 13 ii 373-381. Compare Bruni and Finzi this vol. ii 2 ; Bruni and Padoa Abstr. 1908 i 343).-The freezing-point surface and the course of the isotherms of solutions of mixed optical isomerides give ZL clue to the molecular condition of the dissolved substances and indicate the existence or non-existence of racemic molecules in the liquid. On adding to a p-xylem or ethylene bromide solution of methyl diacetyl- tartrate increasing quantities of methyl diacetylracemate it is found that the freezing-point depressions are always less than the normal values; the molecular weights are hence abnormally great so that there must be a certain number of undissociated racemic molecules in solution. The abnormality diminishes as the concentration of the racemic compound increases which must be attributed to the fact that the excess of the diacetyltartrate undergoes a continuous relative decrease its influence in retarding the dissociation thus becoming less.If temperatures are bken as ordinates and total concentrations as nbscisw the freezing points of the solutioiis of methyl diacetyltartrate and diacetylracemate and the cryohydrate point of the mixed solutions lie on a straight line if the racernic compound undergoes complete VOL. LXXXVIII. ii. 670 ABSTRACTS OF CHEMICAL PAPERS. dissociation in the solution ; where there is partial racemisation this is no longer the case. The three dimeiisional curves obtained with the concentrations of the d- and I-compounds and the temperature as co-ordinates also have different forms for the cases where t>here is partial raceinisation and complete dissociation.RIethyl diacetylracemate also exists partially non-dissociated in solution in diphenylmethane or naphthalene in the latter solvent a t temperatures as high as 80”. The freezing-point constant for diphenyl- methane is 71 and not 67 as was stated by Eijkman (Abstr. 1890 324). Ethyl r-dibrornophenylpropionate is partially non-dissociated in naphthalene solution in presence of the rl-cornpound. T. H. P. Possibility of Resolving Racemic Compounds by Circularly Polarised Light. The Primary Production of Optically Active Substances. ALFRED RYK (Zeit. physiknl. Chenz. 1904,49,641-687 ; Ber. 1904 37 4696-4700).-As a starting point Cotton’s experi- ment is taken according to which the d- and I-components of Fehling’s solution possess different coefticients of optical absorption for circularly polarised light of definite sign.It can also be shown that the positions of the maxima of optical absorption and of photochemical sensitiveness coincide and the conclusion is drawn that the extent of decomposition of these sensitive compounds in circularly polarised light is not the same for each. On the strength of Ostwald’s work on the independelit optical absorption of the ions in dilute solution the author drams the further conclusion that this difference of sensitiveness persists in the inactive Fehling solution and must therefore result in this solution gradually becoming active under the decompobing influence of circularly polarisecl light.For the details of each step in this indirect proof reference must be made to the original. The extensive occurrence of optically active compounds in nature may thus be referred to the production in excess of one kind of circu- larly polarised light at the earth’s surface. This result is reached by the reflection of the plane polarised rays of sunlight from the enrface of the sea assisted by the influence of the earth’s magnetism. Incidentally it was found that the same effect was produced by cl- and by I-circularly polarised light on sensitive optically active photo,ar,iphic preparations (silver tartrate paper and silver bromide plates sensitised with chlorophyll). The progressive dissociation with dilution of a racemic salt was followed in the case of Fehling’s solution and it WRS shown that the production of the racemic salt obtained by mixing tl- and I-Fehling’s solutions is accompanied by a change of colour.J. C. P. Formation of Salts in Solution especially in the Case of Substances Exhibiting Tau tomerism (Pseudo-acids Pseudo- bases). I. JULIUS W. BRUHL and HEINRICH SCHRODER (Zeit. physikal. C‘hem. 1904 50 1-42).-Pull details are given of the spectrochemical investigation of the esters of camphorcarboxylic acid and the sodium derivatives formed from these. The chief conclusions arrived at have already been reported (Abstr. 1904 i 646 and 969). As a further aid secondary result of the investigation following from observationsGENERAL AND PHYSICAL CHEMISTRY. 7 1 made on the variation of the specific refraction of sodium in methyl and ethyl alcohols and its constancy in amyl alcohol (see Abstr.1904 i 969) it seems that the value of the specific refraction for the sodium ion is 12-15 per cent. less than the value for sodium in an unionised compound. J. C. P. Luminescope for Comparing Substances under the Influ- ence of Radium Rays. C. S. STANFORD WEBSTER (J. Soc. Chem Ind. 1904 23 1185-1 186).-The instrument has been designed to facilitate the comparison of different substances in regard to their ‘‘ glow ” and scintillating power under the influence of radium rays. It resembles the spinthariscope but the draw tube with the system of lenses is mounted on one side of a circular metal box which contains a metal disc carrying four screens. By turning a handle the disc can be rotated and the screens brought into position under the observation tube as required.One of the screens is temporarily fixed by wire-spring clips and can be replaced a t will. H. M. D. Induced Radioactivity. KARL A. HOFNANN L. GONDER and VALENTIN W~LFL (Ann. Phpsik 1904 [iv] 15 615-632. Compare Abst4r. 1902 ii 261 397; 1903 ii 402).-Although a uranium salt can be separated into a more and a less active portion the two por- tions finally exhibit the same activity (compare Becquerel Abstr. 1902 ii lli’) so that uranium is itself a radioactive element. When small quantities of salts of other metals are dissolved in concentrated solutions of uranium nitrate and after a few weeks are separated from the uranium by chemical methods the precipitated sulphate sul- phide or oxalate as the case may be is found to have an induced radioactivity which gradually disappears.The degree of activity of the precipitate depends on the nature of the inactive metal it contains ; thus the activity transmitted to lead and bismuth sulph’des is much greater than that transmitted under similar conditions to the sulphides of palladium platinum and mercury. When two inactive metals acquire an induced activity from the same uranium solution their relative activity is independent of the order in which they have been separated from the solution. When the metals are arranged according to the relative activity induced under the above conditions the order is roughly the same whether uranium or radium is used as the inducing agent but is quite different when radiolead or radiobismuth is the inducing agent.For iridium palladium rhodium and platinum are distinguished by the marked activity which is induced in them by radiolead. The chloride iii each case was kept f o r three weeks in common solution with raclioleacl chloride and then the metal was pre- cipitated by formaldehyde o r hyclroxylamine. The a-activity thus induccd 011 these metals is removed on ignition at a bright red heat but the P-activity is oiily slightly diminished by that treatment. This observation leads to the suggestiou that the a-activity is occluded but in conflict with this view i t has been shown for platinum that the activity of the metal is not greater than that of the sulphide obtained by precipitation from the same radiolead solution.At the ordinary temperature the induced P-activity disappears more rapidly than the 6-272 ABSTRACTS OF CHEMICAL PAPERS. a-activity. When the metals palladium iridium and silver (platinum and gold in a less degree) in the solid form have been immersed for several days in an active lead chloride solution obtained from pitch- blende and kept a t 60° they exhibit marked radioactivity. When active bismuth salts prepared from pitchblende are used as inducing agents in the manner previously described they lose their own activity and do not regain it i n contrast to the radiolead preparations. The process of induction is supposed to consist in the emission of particles from the primarily active substance those particles adhering then to the atoms and ions which are susceptible to the induction.J. C. P. Action of Redium Rays on Caoutchouc. RUDOLF DITMAR (Chem. Centr. 1904 ii 1652-1653 ; from Gummi-Zeit. 19 3-4. Compare Abstr. 1904 i 757).-The power of radium rays of pene- trating caoutchouc has been examined by Madame Curie’s radiographic method. A ‘‘ coucentrated ” preparation of de Haen’s radium barium bromide was used but the action of this substance was not sufficiently powerful to induce radioactivity in the caoutchouc itself. The follow- ing conclusions have been derived from the results of the experiments of the author and the investigations of Madame Curie. The radium rays (y-rays) penetrate both crude and vulcanised rubber but the former more readily than the latter probably owing to its porosity. Caoutchouc itself may be rendered radioactive (Madame Curie).The loss of act,ivity of caoutchouc in air does not take place in accordance with the law which usually obtains for solids. E. W. W. Variation of the Resistance of Bismuth in a Feeble Magnetic Field. C. CARPINI (A’uovo Cim. 1904 [v] 8 171-1 86).-The variation of the resistance of bismuth in both feeble and strong magnetic fields is a hyperbolic function of the strength of field and is also a sine function of the angle made by the coil of bis- muth with the lines of force No hysteresis phenomena were observed on changing the strengthof field. When an alternating current is em- ployed the change of resistance is far smaller than with the corre- sponding direct current. The curves appear to depend in a very com- plicated manner on the amplitude and period of the alternating cumen t T.H. P. Electric Conductivity of Phosphorus Emanation and of Freshly Prepared Gases. EUGDNE BLOCH (Ann. Chim. Phys. 1905 [viii] 4 25-144. Compare Abstr. 1903 ii 206 ; 1904 ii 11’7).- The electric conductivity of dry air which has passed over phosphorus or oE freshlyprepared gases (such as hydrogen or carbon dioxide ob- tained by the action of acid on metals or marble respectively) is in each ca.se due to the same cause namely the presence of ions which diffsr in their mobility and their coefficient of recombination from the ordinary ions. M. A. W. Electrolytic Conductivity of Colloidal Solutions. G. E. MALE’ITANO (Conipt. rend. 1904 139 1221-1323).-Cnrefully purifiedGENERAL AND PHYSTCAJ CHERITSTRY. 73 colloidal solutions have an electrolytic conductivity greater than t'hat of water ; this is however due to the small quantity of soluble electro- lytes unavoidably present in such solutions and not to the colloidal granules themselves which appear to be without electric charge for the clear liquid obtained after filtering a colloidal solution through a collodion membrane lias the same electrolytic conductivity as the original solution or the residual liquid rich in colloidal granules ; thus a solution of ferric chloride which had been heated to 100' and con- tained 0.371 per cent.of clilorine and 0.132 per cent. of iron had k = 0.02301 the clear colourless filtrate contained 0.367 per cent. of chlorine and 0.077 per cent. of iron and had k=0*02299 and the residual colloidal liquid contained 0.350 per cent.of chlorine 0.182 per cent. of iron and had k = 0.02273 and similar results were obtained with colloidal solutions of arsenic pentasulphide or white of egg. 11. A. W. Conductivities of Certain Electrolytes in Water Methyl and Ethyl Alcohols and Mixtures of these Solventp. Relation between Conductivity and Viscosity. HARRY C JONFS and CHARLES G. CARROLL (Amer. Chem. J. 1904 32 521-583. Compare Jones and Lindsay Abstr. 1903 ii 55)-The conductivity of cadmium iodide bas been determined a t 0' and 25' in water methyl and ethyl alcohols and mixtures of water with each of these alcohols. The determinations were made a t dilutions varying from 16 to 256. The results show that cadmium iodide does not exhibit a minimum in mixtures of methyl alcohol and water at 25' but that a minimum appears a t 0' in a mixture containing 76 per cent.of the alcohol a t concentrations of not less than 64 but disappears a t greater dilutions. In mixtures of ethyl alcohol and water a t 2 5 O no minimum is observed. Determinations of the conductivity of sodiuni iodide a t 0' and 25' in water and in methyl alcohol of 25 50 and 75 per cent. strength have shown that a minimum is exhibited but is more pronounced a t 0' than a t 25". A t 25' a minimum occurs in 75 per cent. methyl alcohol a t a concentration of 32 but beyond this dilution the minim:L are only observed in the 50 per cent. mixture. A t O' the minimum appears in the 50 per cent. mixture only. The conductivity of solutions of calcium nitrate in water methyl alcohol ethyl alcohol and mixtures of each of these alcohols with water was determined a t 0' and 26" but in no case was a minimum observed.Determinations of the conductivity of hydrogen chloride in methyl alcohol of 50 69.75 90 and 100 per cent. strength gas70 results which show that in each case the molecular conductivity decreases as the concentration increases. In the 69.75 per cent. mixture a maximum is obtained both at 0' and 25'. I n the 90 per cent mixture the results are regular and a limiting value is reached at a dilution of 157.67. It is found that hydrochloric acid shows a minimum both in the 69.75 and the 90 per cent. mixtures A t 25' the minimum appears a t a dilution 178.75 and a t 0' it appears at all dilutions.The conductivity of sodium acetate in acetic acid of 25 50 75 and74 ABSTRACTS OF CHEMICAL PAPERS. 100 per cent. strengths was determined at 2 5 O but the results were irregular and led to no definite conclusions. The dissociation of potassium iodide sodium iodide ammonium bromide ammonium iodide and lithium nitrate in 50 per cent. methyl alcohol WAS calculated from conductivity determinations. I n the case of potassium and sodium iodides it was found that the dissociation is greater in the mixture of methyl alcohol and water than in either of the pure solvent; and that the dissociation of potassium bromide is greater in the mixture than in pure water. The dissociation of hydrochloric acid in methyl alcohol of 69.15 per cent. strength is also greater than in water at the corresponding dilution but this is not the case for the 90 per cent.alcohol. The dissociation of potassium and sodium iodides in water methyl alcohol and ethyl alcohol is found to agree quantitatively with the hypothesis of Dutoit and Astoii (Abstr. 1897 ii 456). It is suggested that the greater dissociation found in the 60 per cent. mixture of methyl alcohol and water is due to the presence of a hydrate CH3*OH,3H,O formed by the association of four simple mole- cules whereas in water the number of molecules associated with one another a t the ordinary temperature is less than four. An explanation has been suggested by Jones and Lindsay (Zoc. cit.) t o account for the occurrence of the minimum point in the conductivity values. It is now shown however that t'his explanation is not alto- gether satisfactory and that the decrease in the coiiductivity of electrolytes in binary mixtures of various alcohols and water some- times accompanied by a minimum conductivity is due mainly to a diminution in the fluidity produced on mixing the solvents arid the consequent decrease in ionic mobility.The hypothesis of Kohlrausch (Abstr. 1903 ii 403) that each ion is invested with an atmosphere of the solvent of considerable thickness is shown to be true in the case of binary electrolytes not only in aqueous solutions but also in methyl and ethyl alcohols. The relation between conductivity association and viscosity is discussed and the hypothesis is put forward that the cmductivities of comparable equivalent solutions of binary electrolytes in certain solvents such as methyl and ethyl alcohols arid acetona are inversely proportional to the coefficient of viscosity of the solvent and directly praportional to its association factor.This relation has been tested in a number of cases and found to be true in each. The results of the experiments are tabulated an(l are also plotted as curves. E. G. Relation betwaen the Size of the Molecular Complex and the Temperature-coefflcient of Expansion in Different States of Aggregation. WILHELM VAUBEL (J. pv. Chevz. 1904 [ii] 70 503-508).-As the molecular vibrations and the consequent tempera- ture-coefficients of expansion of a substance in the gaseous and in the liquid states of aggregation must be in direct piLoportion to the size of the molecules it is proposed to calculate the number of gas-molecules associated in the liquid-molecule from the relation to each other of t hc coe ffici en t s .GENERAL AND PHPSICAL CHEMISTRY.75 The results so obtained agree with those from other methods for chlorine bromine phosphorus bismuth water ether acetone and methyl ethyl and propyl alcohols ; in most other cases the results are too high. Contrary to previous determinations of the molecular weights of homologous series of organic conipounds iii the liquid state the value obtained by the present method increases with the vapour density. G. Y. Effect of Mechanical Vibration on Carbon Dioxide near the critical Temperature. ~VALTER P. BRADLEY A. w. BROWNE aucl C. F. HALE (Physical Review 1904 19 25S-272).-When a system containing bot-h liquid and gaseous carbon dioxide between temperatures of 29' and 31.2" is subjected in a Cailletet apparatus t o proper mechanical vibration a peculiar fog effect is produced.This is due not to impurities in the gas but to a change of state in the liquid and vapour phases resulting respect'ively in the formation of minute bnbbles and drops. The change of state is produced by a1tern:ite compression and dilatation effected by the vibrations of a steel rod clamped t o the pressure jacket of the apparatus. Under given conditions of temperature pressure length of mercury column and relative volume of the phases vibrations of a certain definite pitch are required in order to produce the fog effect. The present paper deals with the measurement of variables determining the pitch of the effective vibrations.J. C. P. Regularities in Homologous Series. OTTO BIACH (&it. physikal. C'hem 1904 50 43-64).-1t is well known that in certain homologous series the melting point alternately rises and falls (com- pare also Lnmouroux Abstr. 1899 i 479 ; Henry ibicl. 735) and in the present paper the author attempts to show that this oscillatory variation is characteristic of all physical constants. T.V hen the constants themGelves vary more or less regularly their differences may be found to oscillate or the differences between these first differences may exhibit the required peculiarity. By way of explanation the author makes the supposition that the vttlency of hydrogen is not exactly one-fourth that of carbon.J. C. P. Rate of Chemical Auto-heating (Adiabatic Reaction Kinetics). GE~RGE BREDIG and P. EPSTEIN (Xeit. cmorg. Chem. 1904 42 341-352).-A theoretical paper i n which the following question is discussed. Given a chemically reactive system of the initial concen- tration a and having a t constant temperature the rate constant kYL which is variable wit11 the temperature T,,whilst in the reaction in ques- tion the amount of heit 9 is evolved for each formula-weight g ; it being assumed that the reaction proceeds in a vessel impermeable to heat so that the system is thus heated or cooled by its own heat of reaction. After what period oE time 2 will n definite temperature 2'; and a t the same time a definite chemical transformation x be reached ? To illustrate the equations deduced the action 2H,O = 2H20 + 0 w n s studied under the catalytic influence of potassium iodide and so far as possible under adiabatic conditions.A t a definite time a76 ABSTRACTS OF CHEMICAL PAPERS. mixture of hyclrogen peroxide and potassiuin iodide was introduced into a Dewar-Weinhold flask and the auto-heating of the syst'em estimated after definite intervals of time. A. McK. Some Thermochemical Rules relating to the Possibility and Prognostication of Reactions. MARCELLIN BERTHELOT (Compt. rend. 1904 139 1005-1009).-A theoretical paper not suitable for abstraction. If. A. W. Thermochemical Studies. DAMEL LAGERL~FF (J. pr. Chew. 1904 [ii] 70 621-559. Compare Abstr. 1904 ii 382 605).-The conclusion of the author's mathematical discussion of the theory of the heat of formation of carbon compounds.G. Y. Combustion of Sulphur in the Calorimetric Bomb. H. GIRAN (Compt. vend. 1904 139 1219-1221).-Tho author has measured the heat of combustion of sulphur in oxygen in a calorimetric bomb under varying pressures estimated the sulphur dioxide and sulphur trioxide in the gaseous products and calculated the heat of formation of sulphur dioxide under different pressures; the results show that the heat of formation of sulphur dioxide and the proportion of sulphur converted into the trioxide increase with the pressure and the following table comprises the results Ratio of weight of snlphur converted into trioxide to Heat of Pressure. formation of SO,. total weight of sulphur. 2.5 ) ) 70-43 0.142 5 ) 71.60 0.165 10 ) ? 72'19 ) 0-184 75 ,) 74-45 ) ) 0.188 25 ) ) 77.88 0.228 80 ) 78'41 0.272 35 9 80.26 0'294 40 2 80.88 0.307 1 ntlllo.;.69.80 Cal. (by extrapolation) - 'LO > 75-52 0.219 45 ) ) 81.13 0.312 I t is probable that the increase with the pressure in the heat of formation of sulphur dioxide is due to the formation of tt higher oxide of sulphur as the residue contained in the bomb shows the characteristic properties of the anhydride of persulphuric acid. The heat of formation of sulphur dioxide under atmospheric pressure is 69.80 Cal. M. A. W. Villari's Critical Point in the Case of Nickel. I(. HONDA and s. SHrMrzu (Ann. Physik 1904 [ iv] 15 855-859).-The authors adhere to their former conclusion (Ann. Yhysik 1904 [iv] 14 791) that nickel does not exhibit this critical point.J. C. P. Fusion of Dissociating Compounds and the Degree of Dissociation of the Fused Substance. ROBERT KREMANN (ilrfofiatsh. 1904 25 1215-1269).-From theoretical considerations itGENERAL AND PHYSTCAL CHEMISTRY. 77 is shown that if curves are constructed with melting points as ordinates and molecular percentages of one substance which forms an additive compound with the other as abscissa the more strongly the additive compound dissociates on melting the flatter is the curve near the maximum Tut its melting point. The melting point of the additive compound is affected less by addition of one of the components than of an indifferent substance as the former takes part in the equilibrium between the additive com- pound and its components.The molecular depression of the melting point of the additive cornpound is calculated from the heat of fusion or is determined directly by addition of iudifferent substances unless these act as solvents when the molecular depression produced increases with the dielectric constant. The extent to which the additive compound dissociates on melting can be determined within 1 per cent. by comparison of the theoretical with the experimental curve. The additive compound of aniline with phenol has the molecular depression of the melting point A = 0*278" and when fused dissociates to the extent of 20 per cent. The curve for mixtures of phenol and picric acid has eutectic points a t 36" and SOo for mixtures containing 7 and 59 molecular per cent. respectively of picric acid ; the additive compound has A = O04Ol0 and a dissocia- tion degree of 27 per cent.For mixtures of trinitrotoluene and naphthalene the curve has eutectic points a t 72" and 71" representing 12 and 85 molecular per cent. respectively of trinitrotoluene ; the additive compound melts a t 96.5". The curve for the mixtures of naphthalene and picric acid (compare Saposchnikcff and Rdultowsky Abstr. 1904 i 399) has eutectic points at 78' and 111" for mixtures containing 4 and 92 inolecular per cent. respectively of picric acid; the additive compound melts a t 147'. The mixtures of nitroso- dimethylaniline and aniline show eutectic points a t - 10' and 75" representing 2.5 and 90 molecular per cent. respectively of nitroso- dimethylaniline ; the additive compound contains 33 molecular per cent.of aniline melts a t 93" has A=O*202' and when melted dissociates to the extent of 15 per cent. Influence of Substitution in the Components on the Equilibrium of Binary Solutions. ROBERT KREMANN (Moncctslh. 1904 25 1271-1310. Compare foregoing abstract).-The melting point curve of mixtures of 2 4-dinitrotoluene and naphthalene has eutectic points at 53' and 56" for mixtures containing 28 and 69 molecular per cent. respectively of naphthalene ; the maximum for the additive compound is 59'. The curve for mixtures of nitrotoluene and naphthalene falls to one eutectic point a t 37" for a mixture con- taining 38 molecular per cent. of nitrotoluene. The curve for mix- tures of trinitrobenzene and naphthalene falls to eutectic points at 77" and 114' for mixtures containing 8 and 88 per cent.respectively of naphthalene ; the maximum for the additive compound (Hepp Abstr. 1883 317) lies a t 151". The curve for mixtures of rre-dinitrobenzene and naphthalene falls to eutectic points a t 50.3' and 50.5" for mixtures containing 60 and 42 molecular per cent. of naphthalene; the curve between the eutectic points is extremely flat the maximum G. Y.78 ABSTRACTS OF CHEMICAL PAPERS. the melting of the easily dissociated additive compounds lies at 60%' ; such mixtures are termed by the author $-isomorphic. The curve for mixtures of nitrobenzene and naphthalene falls to one eiitectic point a t - 6.8' for a mixture containing 14 molecular per cent. of naphthalene. The curve for mixtures of o-nitrophenol and naphthalene has one eutectic point a t 303 for a mixture containing 30 molecular per cent.of naphthalene. The curve for mixtures of p-nitrophenol and naphthalene falls to a eutectic point a t 73" for a mixture coiitaiiiing 77 molecular per cent. of naphthalene. The curve for mixtures of s-trinitrobenzene and aniline has a eutectic point at 101' for a mixture containing 18 iiiolecular per cent. of aniline and another a t - 6 " for a mixture containing 0.3-0.7 molecular per cent. of trinitrobenzene ; the maximum a t 125' repre- sents an equimolecular mixture. The curve for mixtures of nz-dinitrobenzene and aniline falls from the melting point of dinitrobenzene to 40" where there is a break and then to a eutectic point a t - 8' for a mixture containing 4 molecular per cent.of nz-dinitrobenzene ; the break a t 40' represents an additive compound containing 20 molecular per cent.. of nz-dinitrobenzene which melts a t 40". Mixtures of nitrobenzene and aniline give a curve which falls to a eutectic point a t - 39.8" for a mixture containing 41 molecular per cent. of nitrobenzene. The curve for mixtures of trinitrotoluene and aniline falls t o eutectic points a t 60' and - 8" for mixtures containing 85 and 1.5 molecular per cent. respectively of trinitrotoluene ; the maximum between the eutectic points is 85'. The curve for mixtures of dinitrotoluene and aniline falls in an almost straight line from the melting point of the former to a eutectic point at - 13' for a mixture containing 11 molecular per cent. of dinitrotoluene.The curve for mixtures of nitrotoluene and aniline has one eutectic point at - 17" for a mixture containing 15 molecular per cent. of nitrotoluene. G. Y. Additive Compounds of Nitrosodimethylaniline. ROBERT KREMAPU" (MonaGtsI~ 1904 25 1311-1329. Compare foregoing abstracts).-Examination of the melting point curves for mixtures of nitrosodimethylaniline with p-toluidine o-toluidine m-xylidine P-naphthylamine and phenol shows that the following additive com- pounds are formed. The additive compound with ptoluidine contains 66.6 per cent. of nit rosoclimethylaniline ; the eutectic points a t 28' and 48.5" represent mixtures containing 27 and 33.4 per cent. of p-toluidine. The additive compound with o-toluidine contains 66 per cent. of nitrosodimethylaniline and melts a t about 70" ; the eutectic points a t 67" and below - 18' represent mixtures containing 26 and over 95 per cent.of o-toluidine. The additive compound with P-naphtliylamine contains 60 per cent. of nitrosodimethylimiline aiid melts at 86" ; the eutect.ic points a t 81' The percentages are in molecular equivalents.GENERAL AND PHYSICAL CHEMISTRY. 79 :md 75" represent mixtures containing 35 and 82 per cent. of nitroso- dimethylaniline. The additive compound with phenol contains 6 6.6 per cent. of nitrosodimethylaniline and melts at 91"; the eutectic points at 24' and 74" represent mixtures containing 21 and 88 per cent. of nitrosodimethylaniline. Two additive compounds are formed with m-xylidine; the one contains 60 per cent. of nitrosodimethylaniline and melts a t 48"; the other contains 25 per cent.of nitrosodimethylaniline and melts a t 26" ; the eutectic points at 46*5" 2 5 O and below - 11" represent mixtiires containing 63 30 and about 2 per cent. of nitrosodimethylaniline. The melting point curre for mixtures of nitrosobenzene and aniline falls to a eutectic point a t - 13" for a mixture containing 19 per cent. of aniline ; no additive compound is formed. G. Y. Polymerisation in the Liquid and Solid States. GEORGE G. LONGINESCU ( A ~ t n . 5%. Univ. Jc~ssy 1904 3 26-34. Compare Ahstr. 1904 ii 1 1 2 ; 1903 ii 531).-Making use of the relationship (T/G.D)2 = ?a conclusions are drawn in regard to the state of aggregation of a number of organic substances containing bromine or iodine of several orgmo-metallic compounds and of various inorganic substances in the solid or liqriid condition.The bromo- and iodo-derivatives and the organo-metallic compounds consist of simple molecules ; thi3 is also the case for the chlorides and bromides of arsenic and antimony for chrom y 1 chloride germanium chloride nickel carbonyl lead t etra- chloride and selenium oxychloride. Other inorganic compounds for example lead chloride lead bromide and cadmium chloride consist of very complex molecules the association factor being greater than 50. The value of C appears to b j a function of the atomic weights of the constituent elements and a method of ascertaining it is described. The ralationship is also applied to the elements and the molecular complexity of 45 elements in the solid state and 13 in the liquid state is deduced. It would appear that carbon silicon lithium and calcium contain more than 200 atoms in the molecule glucinum magnesiuni and strontium more than 100.Sodium and potassium appear to be more highly polymerised in the liquid than in the solid state. Gener- ally speaking in any periodic group the degree of polymerisation decreases as the atomic weight increases. H. 31 D. Distribution of Soluble Substances between Water and Amy1 Alcohol. WALTER HERZ and HERBERT FISCHER (Ber. 1904 3'7 4746-4753).-Measurernent:; are given of the constants of distribution between water and amyl alcohol of several acids and bases. I n the case of the weaker acids (phenol and acetic acid) a constant value is obtained for all concentrations b J t in the case of the stronger acids (succinic oxdic and picric acids) allowance has t o be made for their degree of dissociation a constaiit value for different concentrations being then obtained for the distribution of the undisso- ciated substance With the bases ammonia methylamine and triethyl- amine the constant is independent of the concentration.The distribution ratio of iodine between the two solvents is constant80 ABSTRACTS OF CHEMICAL PAPERS. for all concenti*ations. But with nmyl alcohol and aqueous potass- ium iodide very different values are obtained for different concen- trations the ratio being also dependent on the nature of the potassium iodide solution. W. A. D. Determination of Molecular Weights in Solid Solutions. FRIEDRICH X7. KUSTER [with WALTER WURFEL] (Zed.pluysikal. Chenz. 1904 SO 65-80. Compare Abstr. 1894 ii 274 ; 1895 ii 439),-The freezing point curve for mixtures of p-dichloro- and p-di bromo-benzenes ascends regularly from the freezing point of the former to that of the latter and is slightly convex to the concentration axis. The separating solid in each case is therefore an isomorphous mixture of the two consti- tuents. The composition of the liquid phase in equilibrium with each isomorphous mixture may be deduced from its boiling point for the boiling point curve of mixtures of the two substances is regular and somewhat similar in form to the freezing point curve. The solubility of the isomorphous mixtures in slightly diluted ethyl alcohol has also been determined and these solubility experiments indicate that the molecules of the crystallised compounds are on the average greater than C,H4C1 and C,H,Br2 but smaller than (C,H,CI,) and (C6H4BrJ2.The increase of solubility from the value for pure p-dibromobenzene is nearly proportional t o the percentage of pdichlorobenzene in the isomorphous mixture. Mixtures of 2 4 6-trichlorophenol and 2 4 6-tribromophenol have been similarly investigated. The former compound is djmorphous and the form which primarily separates on solidification is labile a t the ordinary temperature. Tribromophenol does not exhibit dimor- phism and the form in which i t separates on solidification is iso- morphous with that form of trichlorophenol which is stable a t the ordinary temperature. The freezing point of each compound is lowered by the addition of the other and mixed crystals separate from the liquid mixtures; the mixed crystal series however exhibits a gap.The solubility of mixtures of the two substances in dilute methyl alcohol was studied. It is found that whilst the composition of the solutions changes regularly the mixed crystals in equilibrium with the solutions undergo a t one point an abrupt change both in appearance and in composition. When the concentrations of the cldoro- and bromo- compounds in solution are plotted against the molecular concentration of the bromo-compound in the solid phase two straight lincs are obtained for each compound corresponding with the labile and stable forms. The solubility of each form of trichlorophenol is proportional to its concentration in the solid phase the solubility of the labile form being greater than that of the stable form. I n the case of the mixed crystals containing labile tribromophenol the solubility of the latter is proportional to its concentration in the solid phase but on the other hand the solubility of the stable form of tribromophenol increases as its concentration in the solid phase diminishes. This peculiarity is pro- bably due to an associating tendency on the part of tribromophenol and this view is supported by determinations of the molecular weight in alcoholic solution. J. C. P.INORGANIC CHEMISTRY. 81 Stoichiometrical Lawa end the Atomic Theory. LOUIS HENRY (BUZZ. Acad. Roy. Relg. 1904 975-983).-The author dis- agrees with Ostwald’s view that since the stoichiometrical laws are deducible. from the principles of chemical dynamics the atomic theory has become unnecessary for this purpose (Trans. 1904 85 508). T. A. H. Lecture Experiments. [Liquid Ozone. Solid Nitrogen. J HUGO ERDMANN ( B e y . 1904 3’7 4739-4744).-An apparatus is described for storing and manipulating a solution of liquid ozone in liquid air also another apparatus used f o r the preparation of solid nitrogen by kapidly boiling off the liquid gas and suitable for demonstra- tion on the lecture table. E. F. A. Bunsen Burner with Sieve Attachment. FELIX A LLIHN (Chent. Zed. 1905 29 34).-A piece of gauze is attached to the top of the ordinary Marshall burner whereby a flame is obtained which is uniformly hot a i d especially suitable f o r heat,ing ylr?.tinum vessels. A. McK.