5% of wool fat stable sols containing 6 mg. per c.c. are formed. N. H. HARTSHORNE.

Polychrome mercury hydrosols. R. FEICK (Kolloid-Z., 1925, 37, 257-267).-The theories of the colours exhibited by metal sols are discussed, prominence being given to that of Mie (Ann. Physik, 1908, 25, 377), in which the colours are stated to be dependent on the optical constants of the dispersed substance (in the massive state) and the degree of dispersion. By the reduction of mercurous nitrate solutions with pyrocatechol, sodium hyposulphite, or quinol, in the presence of gelatin and varying amounts of nuclear sol," a series of mercury sols of different degrees of dispersion has been prepared, with colours, viewed by reflected light, varying from brownish-black for the coarser suspensions, through browns, reds, and greens to indigo for the most disperse sols. The particle size for each sol was calculated on the basis of Mie's theory and varied from 300 to 80 μp. Within this range, the theory appears to hold satisfactorily for mercury. In a short discussion of the application of the theories of colloidal colours to photohalide effects, it is concluded that the darkening, e.g., of silver chloride, is due to the formation of colloidal silver, and not of a subchloride. N. H. HARTSHORNE.

General colloid chemistry. XVII. Analysis and constitution of colloidal gold. III. L. FUCHS and W. PAULI (Koll. Chem. Beihefte, 1925, 21, 195-240).-The study of a large number of gold sols has shown that progressive dialysis is accompanied by the formation of hydrogen ions. The concentration of these is the same for sols prepared by reduction both by formaldehyde and by hydrogen, namely, about 1 x 10-5 N, measured by microconductometric titration with barium hydroxide. Hydrogen ions derived from carbonic acid in the air, silicic acid from the glass vessels, and organic acids from the parchment paper are shown to be quite insufficient to account for the phenomenon. Further, no other cations can be found in appreciable quanti

ties in the sols. It is therefore concluded that the ions are the partners of the gold particles which are considered to be covered by a layer of the anions of a gold acid. From the titration data and from measurements of the particle number and particle size, it is possible to obtain values for the charge number of the particles, but it is pointed out that the presence of particles invisible under the ultramicroscope may, on account of their relatively large surface, give a quite fictitious value to the result. Bearing this point in mind, the connexion between flocculation and charge values is discussed.

N. H. HARTSHORNE.

Plant colloids. XVI. Behaviour of starch components towards iodine, and their protective colloid action. M. SAMEC and R. KLEMEN (Koll. Chem. Beihefte, 1925, 21, 55-77).-The colour of starch-iodine complexes is to a large extent independent of the average particle size of the colloid and is not influenced by pairing with phosphoric acid, or substances containing it, or with silicic acid, or

by cations bound to amylophosphoric acid. There is no simple relationship between the iodine colour and the protective action towards gold sols. The quantity of iodine taken up by aqueous solutions of single starch components is different for single grain ingredients. With equal potassium iodide concentrations amylo-substances take up more iodine than erythro-substances. The iodine adsorption of amylosubstances increases with increasing potassium iodide concentration up to a maximum, and up to the point of visible coagulation is not appreciably altered by ageing, but falls when the solution is heated to 120°; increase in the degree of dispersion increases the adsorption, and vice versa. Amylopectin adsorbs less the longer the period of boiling necessary for the isolation of the sol. The adsorptive power of erythroamyloses falls with increasing particle size; it is appreciably raised by the presence of potassium iodide, and in contrast with the amyloamyloses excess of the salt does not cause precipitation of the iodine complex. Soluble starch and dextrin generally adsorb less iodine the lower the average particle size. The amyloamyloses have a better protective action towards gold than the erythroamyloses. Generally the protective action increases with increasing dispersity. N. H. HARTSHORNE.

Physical properties of serum on addition of "water-binding" substances. R. FÜRTH and R. PECHHOLD (Biochem. Z., 1925, 164, 9-17).See A., 1925, ii, 1056.

J. KABELIK (Kolloid

Nephelometry of serum. Z., 1925, 37, 274-283).—A review of work on the nephelometry of animal and human sera, carried out by the author and his collaborators [cf. Kabelik and Lednicky, Biol. Listy, 1922, 8, 212; Kabelik, ibid., 1924, 10; Centr. Bakt. Ref., 1925, 78, 185; Loutoucky, ibid. (in preparation); Lednicky, Kolloid-Z., 1923, 32, 12; Stošek, Spisy lék. fak. v. Brně, 1922/23, 18, A 8; Zak, ibid., 1923/24, II 2, A 12; Trapl, ibid., 1923/24, II 9].

N. H. HARTSHORNE.

Colloidal properties of sericin. G. A. BROSSA (Atti R. Accad. Sci. Torino, 1925, 60, 231-240).— Extraction of silk cocoons with water in an autoclave for 1-2 hrs. at 1 atm. caused a loss of weight equal to 20-22% of that of the cocoons and the production of a turbid, green or yellow liquid. On keeping, flocculation of the liquid occurred. This is promoted by traces of acid, the optimum hydrogen-ion concentration being РH 4.4 4.6. The precipitate is peptised by heating. By evaporating the extract a product consisting mainly of proteins was obtained. By subjecting the original extract to electro-osmosis an almost colourless suspension was obtained. The clear liquid remaining after the particles were removed by centrifuging gave protein reactions, but, unlike the original extract, did not flocculate spontaneously or on addition of acetic acid. The properties of these two protein products are compared with those of the a- and B-products obtained by Anderlini by the alcoholic extraction of silk cocoons. F. G. TRYHORN.

Action of electrolytes on the stability of disperse sulphur solutions prepared by the mechanical method. P. P. VON WEIMARN and S. UTZINO (Mem. Coll. Sci. Kyōtō, 1925, A, 8, 291306). Colloidal sulphur solutions have been prepared by grinding sulphur with grape-sugar and treating with water. The suspension carries a negative charge. Stability curves for this sol have been obtained by plotting the concentrations of different electrolytes added, as abscissæ, against the time required for coagulation as ordinates. With sodium, calcium, barium, and cerium chlorides, barium iodide, and hydrochloric acid, the stability rises to a maximum value with increasing concentration of electrolyte, and then decreases to a value considerably below that for the pure sol. The curves for potassium and calcium thiocyanates show two maxima, which is possibly the case for potassium sulphate, but less probable for sulphuric acid. The potassium nitrate curve has no maximum, but indicates instead a continuous decrease in stability with increasing concentration of salt. Salts of tervalent metals display a greater activity than those of bivalent metals, since the first maximum is reached at a lower concentration of the former than

of the latter, and has also a higher value. Similarly,

salts of bivalent metals are more active than those

of univalent metals, and acids are also very active. The coagulating power of an electrolyte appears to be related to its ability to form hydrates, and the double maxima in the case of salts containing sulphur are ascribed to "homo-chemical" action between this combined sulphur and the dispersoid. The results obtained are in agreement with von Weimarn's views on dispersion and aggregation (cf. A., 1925, ii, 969).

M. S. BURR.

Influence of electrolytes on glycogen sol, and the origin and inversion of the Hofmeister ion series. S. DOKAN (Kolloid-Z., 1925, 37, 283-296). -The action of electrolytes on glycogen sol increases the sensitiveness of the sol towards dehydrating agents such as alcohol and tannin. The action of different electrolytes under different conditions leads to very varied results. Occasionally only the positive ions (i.e., the ions charged oppositely to the colloid) are effective and their action depends almost entirely on their valency. Under other conditions, both ions appear to come into play, and they act in the order of the well-known lyotrope series, but sometimes in one direction, sometimes in the other. This behaviour is explained as follows. Some of the ions are adsorbed by the colloid and confer on it their tendency to hydrate, whilst the others remaining free tend to dehydrate the colloid. With small ion concentrations, the hydrating action of the ions adsorbed by the colloid exceeds the dehydrating action of the free ions, but with high concentrations the effect is reversed. In the latter case, the dehydrating action of both negative and positive ions is concerned. The total effect on the state of the colloid is the resultant of these factors combined with the effect on the charge of the colloid particles of the adsorbed, oppositely charged ions. The dehydrating action of a substance such as alcohol or tannin affects

the colloid first and then the free ions. With alcohol, both these effects may be observed; with tannin, only the former. The explanation may be extended to meet the case of albumin sol. N. H. HARTSHORNE.

Effect of gum arabic and other emulsifiers on the acid hydrolysis of esters in heterogeneous systems. R. C. SMITH (J.C.S., 1925, 127, 2602-2605; cf. Goldschmidt and Messerschmitt, A., 1900, ii, 200; Callow, A., 1916, ii, 94).-Solid insoluble emulsifiers such as kaolin have very little effect on the rate of hydrolysis of ethyl acetate by hydrochloric acid at 25°. Gum dammar and gum tragacanth increase the rate of reaction in the heterogeneous system, but reduce it in the homogeneous system, which indicates that the effect is not due to catalysis. The probable explanation is that the gum produces a change in the partition ratio of the ester between the phases. Similar results are given by R. CUTHILL.

other esters.

Effects of age on soap solutions. (MISS) R. M. COBB (Ind. Eng. Chem., 1925, 17, 1134-1135).— See B., 1925, 998.

Gel of protein type (found) in the kieselguhr stratum of the Lüneburger Heide. F. V. VON HAHN (Kolloid-Z., 1925, 37, 300-303).-A gel with protein properties, to which the name "cornuit" is given, has been found in the kieselguhr of Neu-Ohe. The substance contains about 2.8% of solid and 0.07% of ash, chiefly diatomaceous residue. It shows many of the properties of gelatin. It may be peptised by sodium hydroxide solution and gelatinised again on neutralisation. The jelly so obtained may be used as a medium for Liesegang ring phenomena. The protective action of the sol is 30 times that of gelatin. Boiling lowers the viscosity, but does not cause coagulation. N. H. HARTSHORNE.

[Behaviour of] colloids under an alternating current. O. BLÜH (Kolloid-Z., 1925, 37, 267-270). -A cell is described in which the behaviour of a sol under the influence of an alternating current may be observed ultramicroscopically. In this, the walls are so close together that electro-osmotic oscillation of the liquid is excluded. In experiments with an old vanadium pentoxide sol, it has been found that some of the rod-like particles attach themselves by one end to the walls of the cell and that the application of the current causes the free ends to oscillate. By increasing the applied potential, a point is reached for any particular particle where its amplitude cannot be further increased, since the wall of the cell prevents further movement. This maximum amplitude is equal to the length of the particle, which may thus be measured. N. H. HARTSHORNE.

Influence of adsorption on the colour of sols and of precipitates. N. R. DHAR (J. Physical Chem., 1925, 29, 1394-1399).-The colours of sols and of freshly-coagulated precipitates are discussed in relation to previous work. They depend on the nature of the material adsorbed. Positively and negatively charged sols of the same substance often differ from each other in colour and from the electrolyte used in the preparation. Yellow silver chromate sol is negative, whilst the red sol is positive (see also Sen and Dhar, Kolloid-Z., 1924, 34, 270). Negatively charged manganese dioxide sol prepared from permanganate and hydrogen peroxide is deep brown in colour. Addition of a small amount of

ferric chloride causes charge reversal and the positive sol is distinctly red. Analysis shows the presence of ferric iron in the coagulum. The colour of hydrated manganese dioxide (prepared from permanganate and manganese sulphate) depends largely on the substance adsorbed during formation of the precipitate. Precipitated in presence of the nitrates of silver, bismuth, mercury, or lead, it is deep black, in the presence of ferric chloride or sulphate, red, and when no electrolyte is added, reddish-brown. Other effects observed with this substance are quoted. The colour of ferric hydroxide precipitated in the presence of potassium dichromate or permanganate or sodium thiosulphate is different from that obtained in their absence. The blue colour of adsorption compounds of iodine with starch, dextrin, basic lanthanum acetate, etc. is due probably to the existence of iodine as the dispersed phase in these media. The colour of complex copper solutions containing alkali and glycerol or sugar is also discussed.

L. S. THEOBALD.

Behaviour of silica gel towards certain alkalis and salts in aqueous solution. W. A. PATRICK and E. H. BARCLAY (J. Physical Chem., 1925, 29, 1400-1405). The removal of sodium hydroxide from aqueous solution by silica gel and the subsequent replacement of the "adsorbed " sodium ions by those of silver, copper, and iron (ferric) has been studied. Known weights of gel of definite watercontent, between the limits 8% and 11%, were shaken at 20° with solutions of sodium hydroxide of concentrations varying from 0-0127 to 0.1325 mol. per litre. The adsorption of alkali (millimol. per g. of gel) can be represented by K.cn, in which K = 0.0171, n = 2.56, and c is the concentration of the solution. The gel containing adsorbed sodium ions was then treated with 0.1M-silver nitrate, 0·1036Mcopper nitrate, and 0.178M-ferric sulphate solutions. in separate experiments. The sodium ions replaced stoicheiometrically. With silver and copper solutions, no adsorption occurs with pure silica gel, but with iron, a slight effect was noted.

are

It is suggested that three types of adsorption should be recognised, viz., (1) chemical, (2) the formation of a molecular layer, and (3) capillary adsorption. With sodium hydroxide, the phenomena partake of the nature of both chemical combination and adsorption, the retention of alkali by the gel being due to chemical forces. The increase in the ratio of alkali taken up to silica with increasing

concentration of the former may be due to the peptisation of the gel by alkali and increasing union with the smaller silica particles thus resulting, or the thickness of the alkali" layer" may increase with the amount of alkali in the surrounding solution. If the silica were of true molecular dimensions, the formation of a definite silicate would be expected. L. S. THEOBALD.

Effect of temperature of formation on the physical character of hydrous aluminium oxide. J. H. YOE (J. Physical Chem., 1925, 29, 14191422). The adsorptive power of hydrated aluminium oxide for arsenite ions has been shown (A., 1925, ii, 107) to be less the higher the temperature of precipitation, an effect which may be due to the formation of a more compact precipitate at the higher temperature. In order to correlate temperature of formation with size and density of particle, the rates of settling have now been determined. Hydrated aluminium oxide was precipitated from the sulphate at temperatures from 0° to 100° under constant conditions. The velocity of settling, 0.014 cm. per sec. at 20° for the first 50 cm., is, however, approximately the same for all precipitates save that formed at 100° and refluxed for 48 hrs. This last showed a velocity of 0.032 cm./sec. Boiling the precipitate formed at 66° for 62 hrs. did not alter the rate of settling.

The hydrated oxide formed at 0° is very gelatinous, but that formed at higher temperatures is less so, although still voluminous. Contrary to Taylor (A., 1911, ii, 542), aluminium hydroxide, precipitated at 66° by ammonia, and heated to boiling, gives a voluminous, and not a granular, product.

Owing to irregularities in the shape and size of particles of the same specimen, attempts to measure their diameters failed. Staining with 0.01% solution of alizarin was also unsuccessful. L. S. THEOBALD.

Hydrates and hydrogels. VII. Isomeric hydrogels of aluminium hydroxide. VIII. Aluminium hydroxide gel of the formula AlO OH. R. WILLSTÄTTER, H. KRAUT, and O. ERBACHER. IX. Silicic acid. R. WILLSTÄTTER, H. KRAUT, and K. LOBINGER.-See this vol., 34, 35, 36.

Reaction between gaseous methyl ether and hydrogen chloride. J. SHIDEI (Mem. Coll. Sci. Kyōtō, 1925, 9, 97-119).-An investigation of the reaction Me,O+HCl Me,O,HCl previously examined by Friedel (Bull. Soc, chim., 1875, 24, 160), whose results suggested that the amount of hydrochloride formed is a minimum when the reactants are mixed in equal volumes. Pure anhydrous hydrogen chloride and methyl ether were mixed in varying proportions, and the volume changes were observed. The equilibrium constants, K, in terms of partial pressures at 1°, 5, 9°, and 19° under a constant pressure of 0.996 atm., were found to be 1.4906, 1.7224, 2-0837, and 3-1744, respectively. In every case, the equilibrium partial pressure of methyl ether hydrochloride was at a maximum when the reactants were mixed in equal volumes, and decreased gradually with increase of excess of one of the components. On plotting log K against 1/T, a

(compound, AB2, m. p. 66.5°; eutectics both melt at 66.5°); m-hydroxybenzaldehyde : picric acid (compound of undetermined composition, m. p. 90°; eutectics both melt at 86.5°); antipyrine: quinol (two compounds, A,B,, m. p. 130°, and AB2, m. p. 134°; eutectics, 102-5°, 118.5°, and 120.5°).

Type III (compound formed which decomposes below the m. p.): Acetamide: salicylic acid (compound, AB; eutectic, 53°; transition point, 65°); anthracene: picric acid (compound, AB; eutectic, 110°; transition point, 151.8°); p-dinitrobenzene: a-naphthylamine (compound, AB; eutectic, 40°; transition point, 81°); p-dinitrobenzene : B-naphthylamine (compound, AB; eutectic, 87°; transition point, 91.2°); carbazole 2: 4: 6-trinitrotoluene (compound, AB; eutectic, 73-5°; transition point, 140°). C. HOLLINS.

System: water and the nitrates of potassium and ammonium at 25°. K. ANDO (Mem. Coll. Sci. Kyōtō, 1925, 8, 283-286).-The conditions of equilibrium were determined by analysis of the liquid and solid phases. Two series of solid solutions, separated by a considerable gap, are formed. In terms of the formula

100m H2O,xNH,NO,,(100-x)KNO, the data characteristic of the system are: solid potassium nitrate saturated with ammonium salt, x=13.5; solid ammonium nitrate saturated with potassium salt, x=65-0; liquid solution saturated with the two salts, x=' =79.41, m=2.18.

L. L. BIRCUMSHAW.

Equilibrium in the system potassium sulphate, potassium nitrate, water at 25°. R. INOUYE (Mem. Coll. Sci. Kyōtō, 1925, 8, 287-290). -The equilibrium conditions were deduced from analyses of both the liquid and solid phases. The solubilities of potassium sulphate and potassium nitrate are, respectively, 11-98 and 38-19 g. of salt in 100 g. of water. The solution saturated with respect to both salts contains 3.95 g. of potassium sulphate and 25-37 g. of potassium nitrate per 100 g.

of water.

L. L. BIRCUMSHAW.

System silver sulphate aluminium sulphatewater at 30°. R. M. CAVEN and T. C. MITCHELL (J.C.S., 1925, 127, 2550-2551).-Mixtures of solutions of the two salts yield no crystalline compound at 30°. R. CUTHILL.

Equilibrium in the systems aluminium sulphate copper sulphate water and aluminium sulphate ferrous sulphate-water at 25°. V. J. OCCLESHAW (J.C.S., 1925, 127, 2598-2602; cf. Caven and Mitchell, A., 1925, ii, 396).-No double salt is formed in the copper sulphate system. From has been isolated. the other system a double salt, Al2(SO4)3, FeSO4,24H2O, R. CUTHILL.

Equilibrium in systems of the type Al2(SO4)3— M"SO, H2O. II. Aluminium sulphate nickel sulphate-water at 30°. R. M. CAVEN and T. C. MITCHELL (J.C.S., 1925, 127, 2549-2550; cf. A., 1925, ii, 396).-No double salt is formed in solution. R. CUTHILL.

Ternary systems. III. Silver perchlorate, toluene, and water. A. E. HILL and F. W. MILLER (J. Amer. Chem. Soc., 1925, 47, 2702-2712; cf. A., 1922, ii, 555).—Of the three binary systems silver perchlorate-water, water-toluene, and silver perchlorate-toluene, the first has already been investigated, whilst the second is difficult to study on account of the extremely low mutual solubility of the compothe range -73-5° to 75°. At the lower temperature, nents. The third has now been investigated over the solubility of the salt is too small for detection; at 25° the saturated solution contains 50·3% of silver perchlorate and is thus unique for the combination of a highly polar solute with a non-polar solvent. AgCIO,,C,H,, and the solubility falls very rapidly Below 22-6°, the solid phase is the compound

with falling temperature.

The ternary system has been studied from the ternary eutectic at 94° up to 91-75°. In ascertaining the composition of the liquid phases, the salt was method. The data show the existence of seven determined directly and the water by an indirect quintuple points, each of which was characterised, and twenty 4-phase equilibria. In addition to the solubility curves for silver perchlorate, its hydrate, and its toluene compound, there are two binodal curves. One of these is submerged and does not reach any of the two-component axes at any temperature. The other binodal curve shows an abnormal distribution of silver perchlorate between the toluene entirely in the latter up to high concentrations, in and water phases, the salt being present almost spite of its great solubility in toluene. This may be due to complete dissociation in concentrated aqueous solution, but it is probable that compound formation and chemical affinities play a large part in all cases of distribution. The intersection of the two binodal curves at points other than their plait points gives rise to a three-liquid system, which is stable from -24.1° to 90°. A. GEAKE.

System Na,SO-NaCl-MgSO, MgCl, H2O. H. J. ROSE (Trans. Roy. Soc. Canada, 1925, [iii], 19, III, 33). The composition of the solution saturated with respect to thenardite, mirabilite, and astrakanite