57 General and Physical Chemistry. A Method for the production of Coloured Flames. By STSCHEGLAYEW (Zeit. physikal. Chem., 1901,39, 11 1-1 13).-'I'o obtain a flame colonred by a metallic salt, lasting for n considerable time in a steady state, the author suggests a blast of air blown horizontally into the surface layer of a saturated solution of the salt and then forming the air current of a bunsen burner. It is stated that if a motor be eniployed to yield a regular air stream at about 60 mm. pressure, most satisfactory results are obtained. L. M. J. Photographs of Spark Spectra. I. Ultra-violet Spark Spectra of Iron, Cobalt, Nickel, Ruthenium, Rhodium, Pallad- ium, Osmium, Iridium, Platinum, Potassium Chromate, Potassium Permanganate, and Gold. By WALTER E. ADENEY (Trans. Roy.Dublin Soc., 1901, ii, 7, 331--338).-Reference must be made to the photographic reproductions accompanying the paper. I n these, many lines observed by Eder and Valenta, as well as by Exner and Haschek, are absent, probably owing, in the majority of cases, to the different methods of sparking employed. J. C. P. Theory of Fluorescence. By WOLDEMAR VOIGT (Arch. NZer. sci. exact. nat., 1901, [ ii], 6, 352-366).-Fluorescent and phosphor- escent phenomena are due to free, incoherent vibrations within the excited substance. The author discusses the electron theory and from results so far obtained concludes that the molecules of fluorescing substances are capable of existing in two conditions, in which the electrons have different periods of vibration. The change from one state to the other is conditioned by molecular relationships, but an exciting light wave has also the power of aiding or starting the transi- tion.The electrons pass into the new condition with speeds and elongations which are influenced by the motion given t o them by the exciting wave in the old condition and they perform free, incoherent vibrations with a period corresponding with the new condition. The periods of vibration in the two conditions are very different and one is damped to a much greater extent than the other. J. McC. Chemical Effects produced by Radium Radiations. By HENRI BECQUEREL (Compt. rend., 1901, 133, 709--712).-Various chemical ef-fects produced by radiations from radium have been already observed, as, for example, the action on silver gelatino-bromide, or barium platinocyanide, destructive action on the skin, the coloration produced in rock salt, sylvite, and in varieties of glass or porcelain. The author has observed the following additional actions.(1) The transformation of yellow phosphorus into red, which may be brought about by immersing a sealed glass tube containing radium, enveloped in alnminium foil, into a glass vessel containing the phosphorus, the whole being kept in the dark. The transformation does not continue LXXXII. ii. 558 ABSTRACTS OF CREMTCAL PAPERS. after the removal of the radium. (2) The reduction of mercuric chloride by oxalic acid which takes place in the dark if the radium tube be placed in the mixed solutions. (3) Destruction of the germin- ative power of seeds by exposure to the radiation before planting.Mustard and cress seeds were divided into two portions, of which one was exposed for a week or more to the radiations ; none of the seeds so exposed germinated, whilst of the others used for comparison, 80 per cent. germinated. L. M. J. Induced Radioactivity excited by Radium Salts. By P. CURIE and A. DEBIERNE (Compt. rend., 1901,133,931-934. Compare Abstr., 1901, ii, 216,298).-Tho phenomena of induced radioactivity excited by radium salts are more regular, and the activity is more intense, when an aqueous solution is used instead of the solid salt, The intensity of the induced activity is the same for all substances, whatever their chemical. nature, under the same conditions, and is independent of the pressure of the gas surrounding the exciting and excited bodies.The induced radiation, like the exciting radiation, consists of some rays which are de- flected in a magnetic field and some which are not. Other conditions being the same, the intensity of the induced radiation depends on the free space in front of the excited body; if, for example, several copper plates are placed parallel with one another and about 1 rum. apart, they acquire little activity, but if about 30 mm. apart they all become strongly active. The intensity of the radioactivity which can be excited in a given enclosure depends only on the quantity of radium introduced into it in the form of a solution. The glass of the enclosing vessel generally becomes luminous, and the luminosity finally acquired by any part of the vessel is independent of the position of the radium solution.If a radium solution and zinc sulphide are placed in separate flasks connected by a glass tube bent twice at right angles, the sulphide becomes and remains phosphorescent, and a t the same time exhibits radioactivity, the intensity of the activity being independent of the phosphorescence and equal to that which would be acquired by any other substance under the same conditions. C. H. B. Influence of Radioactive Substances on the Luminescence of Gases. By ALEXANDER DE HEMPTINNE (Compt. vend., 1901, 133, 934-935).-Gases subjected to the influence of radioactive substances become luminous under the electric discharge at higher pressures than under normal conditions. I n this respect,, the Becquerel rays, therefore, resemble Rontgen rays; in both cases also the phenomenon is more marked the higher the molecular weight of the gas.Electromotive Efficiency of the Elementary Gases. 11. Note by EMIL BOSE (Zeit. physikal. Chem., 1901, 39, 114).-A note in which the author acknowledges priority of Richarz regarding some points of his work on this subject (Ahstr., 1901, ii, 589). Observations on the Determination of Transport Numbers of the Ions during Electrolysis of their Solutions. The Be- haviour of Diaphragms during the Electrolysis. By WILHELM HITTORF (Arch. Nger. sci. exact. nat., 1901, [ii], 6, 671-688).--The C. H. B. L. M. J.GENERAL AND PHYSICAL CHEMISTRY. 59 transport numbers recently determined do not in all cases agree with the early determinations of the author, and the divergence is greater than can be accounted for by experimental error.This led the author to a n examination of the influence of diaphragms of porous porcelain, fine silk, and animal membrane in the conductivity cell. These diaphragms are used to keep the concentration of the electrolyte in the middle unchanged by preventing diffusion. I n the case of copper salphate, silver nitrate, and the chlorides of potassium, ammonium, sodium, barium, calcium, magnesium, and cadmium, the transport number is the same whether no diaphragm or one of silk or porous por- celain be used. With cadmium chloride, when animal membrane is used, the transport number of the cation is smaller than when no diaphragm is used. The animal membrane has the power of separating the solution into a more and a less concentrated part, and the less concentrated solution goes in the direction of the negative current, in this way leaving the solution round the cathode more dilute than it would be if no diaphragm were used.The use of animal membrane as diaphragm is without influence on the transport numbers of the ions of the chlorides of potassium, ammonium, and sodium. J. McC. Dissociation of certain Acids, Bases, and Salts at Different Temperatures. By HARRY C. JONES and JAMES M. DOUGLAS (Amel.. Chem. J., 1901, 26, 42S--453j.-The substances investigated mere hydrochloric, nitric, and sulphuric acids, potassium hydroxide, chloride, bromide, iodide, nitrate, sulphate and permanganate, sodium nitrate and ammonium nitrate.The temperature coefficient of conductivity increases (1) with dilution for acids, bases, and snlts ; (2) with rise of temperature for salts ; in the case of acids and bases, chnuge of temper- ature has no appreciable effect on the temperature coefficient of con- ductivity. The amount of dissociation in solutions of the above sub- stances, as measured by the conductivity, is independent of the temper- ature. This fact, in conjunction with the observation that the conductivity of the solutions increases with the temperature, shows that rise of temperature affects the velocities of the ions. J. C. P. Effect of Temperature and Moisture on the Emanation of Phosphorus, and a Distinction in the Behaviour of Nuclei and of Ions. By CARL BARUS (Amer. J. Sci., 1901, [iv], 12, 327-346).-A physical paper dealing with the ionisation of air by its passage over phosphorus at various temperatures.J. c1. P. Pressure as Supplement to Temperature in the Phenomenon of Inflammation. By WALTHERE SPRING ( A ~ c h . N6er. sci. exact. nat., 1901, [ii], 6, 257-261).--No combination took place on subject- i n g an intimate mixture of 2 mols. of cupric oxide and 3 mols. of sulphur to a pressure of 10,000 atmospheres. Combination took place violently when the pressure on a mixture of 2 mols. of cuprous oxide and 3 mols. of sulphur rose to 8000 atmospheres; the pressure was increased gradually so that no heating by compression took place. Sulphur dioxide was formed and the residue consisted only of cuprous sulphide ( 2Cu20 + 3s = 2Cu2S -f- SO,).The ignition temperature of this 5-260 ABSTRACTS OF CHEMICAL PAPERS, mixture at the ordinary pressure is about 126', and pressure to the extent of SO00 atmospheres has the effect of lowering this by moro than 1 0 0 O . The ignition temperature of the mixture of cupric oxide and sulphur could not be determined for the sulphur inflamed at 350°, but it must be higher than this. It would appear that the point of inflam- mation is a function of the pressure, and the experirnents are being continued to ascertain if this is quite general. J. McC. Isotherms for Mixtures of Hydrogen Chloride and Ethane. By N. QUINT GZN (Zeit. physika2. Chem., 1901, 38, 14-26).-1so- therms for hydrogen chloride, ethane, and mixtures containing S6*S3, 59.68, 38.33, and 28-59 per cent. of hydrogen chloride respect- ively, were determined at temperatures from 15' to 55' and the critical phenomena investigated.The mixtures behave very similarly to the mixture of nitrous oxide and ethane investigated by Kuenen (Abstr., 1896, ii, lo), and the author uses his results to test the validity of vander Waals' expression in the case of mixtures. Satisfactory agreement between calculated and observed numbers is obtained. L. M. J. Minimum Value of the Total Heat of Combination. By ROBERT DE FORCRAND (Compt. rend., 1901, 1.33, 681-684).-Thc, author has previoiisly enunciated the relation(L + s)/Y'= 30, or, in the! case of a dissociative change in which a gas isproduced, ( A -t S+ q)/T= 30,where q is the heat of combination. I n the case of a gas and a dis-.sociable compound of this gas, therefore, ( L + S)/T = (,L+X+q)/T'' = 30, hence q/(T' - T) = 30, that is, the heat of combination is propor- tional to the elevation of the boiling point, and when T- T is small, L + X+ q (or the total heat of combination, Q) approximates to L + 8., This deduction is tested chiefly by examples of compounds of ammonia, with metallic salts, in which the value T" is but slightly greater than 234*5',the boiling point of ammonia, for each of which also the totall heat of combination is about 7 to 8 C d , the value L + X for ammonia, being calculated as 7.03 Cal. (Abstr., 1901, ii, 372, 594). L. M. J. Determination of the Heat of Dissociation and of Combus - tion of Acetylene, Ethylene, and Methane. By WILLIAM G,, MIXTER (Amer. J. Sci., 1901, [iv], 12, 347-357).--The heat of disso-.ciation of acetylene is found by explosion in a bomb to be 53300 cnl,, (Thomsen, 47770 cal, ; Berthelot,, 51400 cal.) ; the heat of combustionl of acetylene is 313800 cal. (Thomeen, 310050 cal. ; Berthelot, 3157001 cal.). The heat of dissociation of ethylene was found by exploding a, mixture of ethylene and acetylene, and subtracting the thermal effect; due t o the acetylene ; the author's results vary rz good deal, but indi-, cate that ethylene may be more endothermic than has been supposed The heat of combustion of ethylene is 345800 cal. (Thomsen, 3333501 cal. ; Berthelot, 341100 cal.). The heat of dissociation of methane:, determined in the manner described for ethylene, is found to be, - 19000 cal. (Thomsen, - 211iO cal.; Berthelot, - 21500 cal.). J. C. P.GENERAL AND PHYSICAL CHEMISTRY. 61 New Method of representing Heats of Solution. By HENDRIK W. BAKHUIS ROOZEBOOM (Arch. N&r. sci. exact. nat., 1901, [ii], 6, 430-441).-The heat of solution can best be represented diagramatically by referring the concentration of the solution, not as is usually done, to the number of mols. OF solvent per mol. of dissolved substance, but so that the sum of the number of molecules of solvent and dissolved sub- stance is equal to unity (or 100). The advantage gained is that the curve obtained is a complete one and does not run to infinity. Three forms of curve for heat of mixture of liquid components are known : (1) positive heats only; (2) negative heats only; (3) positive and negative heats according to the relative quantities of the components.The heat of solution of solid substances can be easily obtained from these heat of mixture curves if the heat of fusion is known, and from them also can be deduced the theoretical heat of solution, that is, the heat change which occurs when 1 mol. of salt is dissolved in an infinite quantity of its saturated solution, J. McC. Fusion and Crystallisation. The Theory of Tammann. By PIERRE DUHEM (Arch. NSer. sci. exact. ncct., 1901, [ii], 6, 93-102).- According to the Clausius formula dY’/dP = l/E.Y’/L.(v’ - u) for the variation of the point of fusion with the pressure, in which P is the pressure, 1’ the fusion point a t pressure P, L the heat of fusion at pressure P and temperature T, v the specific volume of the crystalline phase a t P and T, v’ the specific volume of the isotropic phase a t P a n d T, and E the mechanical equivaient of heat, since the valiies of 1/E and Y’/L are positive, the sign of d T / d P must be the same as that of (v’ - v).It is probable that for some substances the value of (v’ - v) is positive up to a maximum value of P, then assumes the value 0, and finally becomes negative ; in these cases, the curve of fusion is concave towards the pressure axis on a system of coordinates. Tammann has found that when an isotropic phase is gradually cooled, the tendency to crystallise is small just below the fusion point; it increases to a maximum as the temperature falls and a t low tem- peratures again becomes small. Tammann interprets this by assuming that if the temperature be lowered sufficiently and the pressure kept constant, a second fusiou point is reached, that is, a second tempera- ture a t which the isotropic and the crystalline phases are in equilibrium.The author shows that the phenomena can be better explained by assuming, instead of a curve of second fusion, a line of false equilibrium. Tammann’s view assumes that there will always be a line along which (v’-v)=O, and a line along which L=O. The new view does not necessitate these lines, which, indeed, are in some cases difficult to admit. J. McC. Folding Point Curves in Ternary Systems. By FRANZ A. H. SCHREINEMAKEBS (Arch. Nier. sci. exact. nut., 1901, [ii], 6, 1’70-192. Compare Abstr., 1901, ii, 224, 305, 372, 436, 641).-The author con- tinues the discussion of the vapour pressure of ternary mixtures.A foldiug point indicates a critical solution, as in this point two liquid layers must be identical. The conditions for critical liquids of the first and second order are developed, and it is shown that if on the curve of62 ABSTRACTS OF CHEMICAL PAPERS. a critical liquid of the first order under constant pressure there rests a critical liquid of the second order, then a t the point of contact the tem- perature must be either a maximum or a minimum. The author develops a formula by means of which it can be foretold whether the temperature at which two liquid layers are identical is raised or lowered by addition of a third component. A critical liquid at a given temperature can only be in equilibrium with vapour a t a certain defi- nite pressure, and change of temperature alters, not only the pressure, bnt also the composition of both liquid and vapour.The effect of chmge of pressure is also fully discussed. J. McC. An Equation for Osmotic Pressure in Concentrated Solution. By C. H. WIXD ( r l r d . N6er. sci. exuct. nut., 1901, [ ii], 6, 714-726) -From considerations similar to those employed by van der Waals in the development of the gas equation of condition, the author deduces the equation of condition in concentrated solution, lil’= [LV+ (cc - a’>/V2] [ V - Obg/V], as a more complete statement of van’t Hoff’s law. N is the osmotic pressure, V the total volume of the system, and ci, CL’, b and 8 are constants. This equation differs from that of van der W a d s inasmuch as the pressure correction may be negative (if u’ is greater .than a) and in that the volume correction contains V in the denominator.The equation correctly expresses the results obtained by Ewan (Abstr., 1900, ii, 195) and by By1 (Pvoefschrift, AmtevJccm, 1901). J. McC. Neutral Salts. By KURT ARNDT (Zeit. anorg. Chenz., 1901, 28, 364--370).-!L’he degree of dissociation of 0-1N solutions of the following are : HC1, 0.9 1 ; HNO,, 0.92 ; &H,SO,, 0.58 ; KOH, 0.89 ; NaOH, 0.84. I n solutions of chloride and nitrate of potassium and sodium, there will be about the same quantity of hydrogen and hydroxyl ions produced by hydrolysis. Since sulphuric acid is less dissociated than potassium or sodium hydroxides, in the solution of thc sulpbates theremill be a slight excess of hydroxyl ions.This excess is too small to be detected by indicators, but the influence of potassium sulphate on the catalysis of ethyl acetate is very ditferent from that of chlorides or nitrates and resembles that of alkalis (Arrhenius, Abstr., 1888, i, 340). The inversion of sucrose by hydrochloric acid is increased by addition of chlorides ; the inversion by snlphuric acid is diminished by potassium or sodium sulphate (Spohr, Abstr., 1885, 1181). The small quantity of the hydroxyl ion can be even more sharply detected by its influence on the birotation of dextrose. The influence of sulphates on the rotation of dextrose is similar t o that of weak basesllevy, Abstr., 1895, i, 586 ; Trey, Abstr., 1897, ii, 299). J. McC. Velocity of Solution of Solid Substances.11. By LUDWIK BRUNER and STANISLAW TOLLOCZKO (Zeit. anorg. Chem., 1901, 28, 314-330. Compare Abstr., 1901, ii, lo).-Experiments on the velocity of solution of benzoic acid cast into a cake and rotated in water show that some of the solid is mechanically rubbed off and becomes suspended in the solution. Drucker’s results (Abstr., 1901, ii, 376) are vitiated by this circumstance. From experiments withGENERAL AND PHYSICAL CHEMISTRY. 63 alabaster, it is proved that the rate of rotation of the plate has a great influence on the velocity of solution, but the volume of the liquid used is without influence. Hydrogen ions are without influence on the speed of solution of calcium sulphate. The velocity of solution is dependent on the structure of the solid substance, smooth gypsum crystals being dissolved more slowly than the granular alabaster.J. McC. The Investigation of Complex Compounds. By GUIDO BODLANDER (Chenz. Centy., 1901, ii, 1109-1 11 1. ; from Sonderabdruck ccus der Festschy. xur Feier desksiebxigsten Geburtstccges Richard Bedekind, 153--182).-Complex compounds are often formed in solution by the combination of a sparingly soluble compound with a molecule or an ion of a soluble one. I n some cases, as, for instance, when silver cyanide dissolves in solutions containing cyanogen ions, the quantity of the more insoluble compound which goes into solution is equivalent to that of the soluble compound, whilst in others, the proportions vary with the quantities present. The solution of silver chloride by ammonia or of cuprous chloride by hydrochloric acid is an example of the latter type.I n such cases, the law of mass action may be applied, and an indication of the composition of the dissolved ions may be derived from the solubility of the less soluble in an excess of the soluble component. Assuming, for instance, the formula Cu,Cl,+, t o represent the complex ions in a solution of cuprous chloride in presence of a soluble chloride, then (CuC1)Wl" = Cu,Cl,+,k, and since the quantity of active cuprous chloride is constant, C1n = Cu,Cl,+,k,. This gives no indication, however, of the value of In, for if n= 2 then the formula of the complex ions may be CuCI,, Cu,C14, or Cu,Cl,, &c. The formula of the complex ions may be calculated from measure- ments of the E.Jf.F.between electrodes of the same metal as that in the complex in a concentration cell containing two solutions which must have either the same concentration in respect of the complex and different concentrations of the soluble component or vice uerstl. The E.M.Y. between silver electrodes immersed in solutions containing equal quantities of silver but unequal quantities of ammonia may be calcu- lated from thelaw of mass action. Assuming the formula of the corn- plex ions to be Ag,(NH,),, then k[Ag,(NH,),] = AgTn.(NH,)n for one solution and k[Agm( NH3),Il = Agl"(NH3)ln for the other. Since the concentration of the complex ions is the same in both solutions, Neglecting the difference of potential at the boundary of the two (Ag = [(NH3)1: (NH3)In.solutions, then the E.M.F. is If the concentration of the free ammonia is the same in both solutions but that of the complex ions different, then : E = 0.058 log(Ag :Agl) = 0*058,, log[(NH,), : (NH,)]. [AeTn(NHQ)n]l:[a~.m(NHQ)n] = (Agl :Ag)Tn, and E; = 0,058 log(Ag,: Ag) = 0.058,m log([Ag,,(NH,),,] : [Agm(NH3)n]). Hence, from E and El, rn and n can be calculated. It has been found by this method that silver chloride or silver nitrate in ammoniacal solution contains the ion Ag(WH,),, cuprous oxide in ammoniacal solution the ion Cu(NH,),, and cuprous chloride64 ABSTHACTS OF CHEMICAL PAPE&8. i n solutions of chlorides the ion CuCl, or CuCI,, according to the concentration. E. W. W. Dissolution of Metals. By T. ERICSON-AUREN and WILHELM PALMAER (Zeit.physikul. Chem., 1901, 39, l-lS).--The law of mass action cannot be applied to calculate the velocity of dissolution of zinc in acids, as the values so calculated do not agree with the experimental results. The authors consider that dissolution is purely electrolytic and occurs solely as a result of local currents; to this is due the slow velocity of dissolution of pure metals. On this assumption, an expres- sion is deduced for the rate of dissolution of a metal in acids of any concentration, the expression, however, involving an unknown quantity, the resistance capacity. This may, however, be deduced from one set of determinations and the velocity of dissoluticn under other conditions then calculated. The values so obtained were found to agree well with those determined experimentally.The tempera- ture coefficient between loo and 50° was found t o be in general about 1.5 t o 2 per cent. per degree. This is far smaller than the usual temperature coefficient of a chemical reaction, but is approximately that of the increase of E.M.R., a result in accord with the theoretical views. L. M. J. By KURT ARNDT (Zeit. yhysikal. Chem., 1901, 39, 64--90).-The velocity of decomposition of aqueous solutions of a,mmonium nitrite was deter- mined at temperatures varying from 60' t o SO0, and at concentra- tions varying from 0.6 ' molar ' t o 0.3 ' molar.' It was found that in the solutions of the higher concentration the increase of tem- perature from 60' to 80' caused a n increase in the rate of evolu- tion of nitrogeu from 0.37 c.q.per min. t o 3.2 C.C. per min, It was observed that the addition of small quantities of acid increases to a very great extent the velocity of decomposition, whilst ammonia causes an equally marked decrease. This suggested that the decom- position is really due to interaction between the ammonium nitrite and nitrous acid produced by hydrolytic dissociation. Erom the effect of the addition of sulphuric acid, it was calculated, on this assumption, that the hydrolytic dissociation is about 0.25 per cent. at TO", a value which agreed with that calculated from the effect of the addition of ammonia. Ammonium srilphate increases the velocity, probably owing to the increase of undissociated ammonium nitrite, whilst sodium nitrite, by increasing also the free nitrous acid, causes a more marked increase of the velocity, The addition of other neutral salts causes, as expected, a decrease of the decomposition.That the decomposition is not a simple change is also indicated by the approximate proportion- ality of the velocity t o the third power of the concentration. Velocity of Decomposition of Ammonium Nitrite. L. M. J. Equilibrium between Carbonates and Bicarbonates in Aque- OUB Solution. By FRANK K. CAMEHON and LYMAN J . BRIGCIS (J. Physicul Cherri., 1901, 5,537-555).-Solutions of sodium carbonate or * The term ' molar ' is used by the author to indicate the niolecnlar w i g h t of a substauce in grams per litre.GENERAL AND PHYSICAL CHEMISTRY. 65 of hydrogen sodium carbonate attain a state of equilibrium in which both salts are present, the composition being dependent on the total concentration, the temperature, ai id the pressure of carbon dioxide in the vapour phase.The concentration of the two salts in solution was determined by titration with hydrogen potassium sulphate with (1) phenolphthalein, (2) methyl-orange as indicator. Curves are given which show the percentage present as normal carbonate at different concentrations and temperatures. At all temperatures, the quantity of normal salt rapidly increases with the concentration until a concentre- tion of about 0*4N, when it remains almost constant ; the proportion of normal carbonate also increases with rise of temperature. I n certain solutions, the author considers a maximum in thecurve is indicated, but fuller examination is deferred.Solutions of potassium carbonate gave perfectly analogous results. Calcium carbonate exists in solution :ilmost entirely as the hydrogen salt, but solutions of magnesium carbonate may contain 50 per cent. of the normal salt. In both cases, the equilibrium and total solubility are greatly affected by the pressure of the carbon dioxide. L. M. J. Precipitation of Colloids by Electrolytes. By WILLIS R. WHITNEY and J. E. OBER (J. Arne?.. Chern. Xoc., 1901, 23, 842-863). -When 30 C.C. of a 1 per cent. solution of barium chloride were added to 200 C.C. of a 1 per cent. colloidal arsenious sulphide solution, complete precipitation of the arsenious sulphide immediately occurred ; it was found that the precipitate contained 0.0152 gram of barium and that an equivalent amount of hydrogen chloride had been pro- duced.By employing an arsenious sulphide solution of half the above strength, it was shown that the composition of the precipitated colloid is independent both of the concentration of its solution and of t h a t of the barium salt. Experiments in which the chlorides of calcium, strontium, and potassium were used showed that the precipitated colloid contained the metals, barium, strontium, calcium, or potassium in the pxoportions of their equivalent weights ; this result supports Whetham’s hypothesis (Abstr., 1900, ii, 62). An index to the literature of colloids is appended. The Standard for Atomic Weights. By THEODOR W. RICHARDS (Zeit. cmoyy. Chenh., 1901, 28, 355-360. Compare Abstr., 1901, ii, 231, 379).-The author supports the proposal of the International Commission to take as standard 0 = 16.On pedagogic grounds, objection cannot be taken to this if, in the development of Avogadro’s rule, use is made of the densities (experimental) of the gases, that is, the actual weights of 1 litre of the various gases at O”, instead of “ specific gravities.” J. McC. By S. H. HARRIS (J. Physical Chem., 1901, 5, 577--586).-The author shows sundry connections between the atomic weights of elements in cliff erent series and calculates the atomic weights of a number of unknown elements to fill the blank spaces in the periodic table. By L. E. 0. DE V~SSER (Rec. Frccv. Chim., 1901, [ ii], 20, 388--393).-Stas has frequently etuphasised the E. G. Mathematical Expression of the Periodic Law. L.M. J. Purification of Gases.66 ABSTRACTS OF CHEMICAL PAPERS. incomplete purification of gases effected by passing them through tubes containing absorbent solids or liquids ; a complete purification can, however, easily be obtained by passing the gas first through a layer of cotton wool which has previously been impregnated with a solution of the absorbent solid and dried in the air and subsequently through closely packed, pure cotton. If the gas attacks the latter, asbestos or fine-threaded glass wool may be used. I n this way, carbon dioxide, generated from marble and hydrochloric acid, can be entirely freed from hydrogen chloride, although Stas has shown the latter to be present in the gas purified by passage through aqueous and solid sodium hydrogen carbonate. W. A. D. A New Method of Manipulating Liquefied Gases in Sealed Tubes. By HENRI MOISSAN (Compt. rend., 1901, 133, 768-771).- When a current of nir at 18' is passed through a mixture of solid carbon dioxide and ethyl or methyl alcohol, the temperature obtained is constantly - 8 5 O ; with methyl chloride or aldehyde, - 90' ; with ethyl acetate, - 9 5 O , and with acetone, - 98'. If the current of air is previously cooled to - 80', the temperature obtained with the solid dioxide and acetone is - 110'. For lower temperatures, recourse must be had to liquid air or liquid oxygen. When liquefied gases have to be sealed up in glass tubes, the opera- tion is greatly simplified by first cooling the tube, so that the gas becomes solid. For EL pressure of 200 atmos., the tube should be of 10 mm. external and 6 mm. internal diameter ; for higher pressures, 7 mm. external and 3 mm. internal diameter ; and for pressures as high as 300 atmos., 6 mm. external and 1.5 mm. internal diameter. The method is applicable when the liquefied gas is to act on some other substance, and if, after the reaction is finished, the tube is again strongly cooled before being opened, the products of the reaction can be distilled off fractionally. The method is not, however, applicable to reactions in which hydrogen is liberated. The author calls atten- tion to the importance of allowing glass tubes which have been strongly cooled to return very slowly to the ordinary temperature. C. H. B.