para-acid more closely corresponds with the phenols than the orthoacid does.

Glyceryl benzoate was prepared by the action of glycerol and soda on benzoic chloride. It forms long, silky needles melting at 72-4°. Mannityl benzoate forms colourless scales melting at 124-125°.

L. T. T.

Boiling Points and Specific Volumes of the Normal Fatty Ethers. By P. DOBRINER (Annalen, 243, 1-22).—The author's experiments lead him to the following results. In the case of metameric ethers, the boiling point falls as the difference in the number of carbon-atoms in the two alcohol radicles diminishes. Methyl ethers have the highest boiling points. In a series of homologous ethers containing one alcohol radicle in common, the difference between the boiling points diminishes as the carbon-atoms increase in number. Methyl ethers form an exception to this rule, as the difference between the boiling points of the methyl and ethyl ethers in such an homologous series is smaller than the difference between the ethyl and propyl ethers.

The boiling points of amyl and hexyl ethers can be calculated approximately, as the difference between the boiling points of homologous amyl and hexyl ethers is equal to one-third of the difference between butyl and heptyl ethers. The difference between butyl and amyl ethers is a little larger, and the difference between hexyl and heptyl ethers is a little less.

The difference between the boiling points of methyl and ethyl compounds is smaller than the difference between ethyl and propyl compounds when the alcohol radicles are directly attached to an oxygen-atom (for instance, in the case of the methyl, ethyl, and propyl salts of the adipic acids), but the case is reversed if the alcohol radicles are directly united to carbon-atoms.

Specific Gravity.-In the case of metameric ethers, the highest specific gravity is exhibited by the ether with the highest boiling point, and the lowest specific gravity by that of the lowest boiling point. In an homologous series, the specific gravity at 0° increases with the carbon-atom, but the specific gravity at the boiling point diminishes as the compounds increase in carbon.

Specific Volume.-The specific volume of an ether is approximately identical with that of an ethereal salt, the radicles of which contain the same number of carbon atoms. The specific volumes of the ethers are larger than the specific volumes of the metameric alcohols. In a series of metameric ethers, the methyl ether has the smallest coefficient of expansion. W. C. W.

Specific Volumes of Normal Alcoholic Iodides. By P. DOBRINER (Annalen, 243, 23-31).-The specific volumes of the following iodides were determined:

The specific volume of an iodide CnH2n+ I is the same as that of an acid of the formula CnH2n+1'COOH. W. C. W.

Boiling Points and Specific Volumes of Phenols and their Ethers. By J. PINETTE (Annalen, 243, 32-63).-In homologous ethers of the same phenol, the difference in the boiling points of the methyl and ethyl ethers is about 5° less than the difference between those of the ethyl and propyl ethers. In ascending the series from the ethyl ethers, the difference gradually diminishes. The difference between phenol and thymol ethers of the same alcohol radicle decreases with an increase of carbon in the alcohol radicle. Meta- and para-cresol have almost the same boiling point, this is also true of their metameric ethers. Orthocresol boils about 11°, and its ethers about 6° lower than the corresponding meta- and para-compounds. The boiling points of propyl, butyl, and octyl phenyl ethers are almost identical with the boiling points of the metameric ethers of meta- and paracresol, and are higher than the orthocresol and thymol-derivatives.

The difference between the boiling point of a phenol and its methylic ether decreases with the increasing molecular weight of the phenol. Specific Volume.-Phenols have a smaller specific volume than the metameric ethers. The specific volumes of phenol and its ethers coincide with those of butyl alcohol and its ethers. Meta- and paracresols and their ethers have approximately the same specific volumes; they are about 1 per cent. higher than the corresponding orthocresolderivatives. The specific volumes of the thymol ethers are relatively small.

The author obtains the following values for the specific volumes of the isomeric xylenes :

Boiling Points and Specific Volumes. By W. LOSSEN (Annalen, 243, 64-103).-The author has compared the boiling points of a large number of organic compounds, and agrees with Dobriner (preceding page) that in an homologous series, the first difference (that is the difference between the boiling points of methyl and ethyl compounds) is less than the second (namely, between ethyl and propyl) when the alcohol radicle is attached to the rest of the molecule by an oxygen-atom. When the radicle is attached to a carbon-atom, the reverse is true in the majority of cases. In many cases, the first and second differences are equal, and in a few cases the first difference is a little less than the second.

The author agrees with Horstmann (Abstr., 1886, 759) that a comparison of molecular volumes determined at the same temperature possesses many advantages not shared by specific volume determinations at the boiling point. W. C. W.

Estimation of Vapour-densities. By C. SCHALL (Ber., 21, 100-101). Further modifications in the apparatus lately described by the author (Abstr., 1887, 695, 882).

J. W. L.

Determination of Vapour-density at Low Pressure. By H. MALFATTI and P. SCHOOP (Zeit. physikal. Chem., 1, 159-164).—In spite of the many modifications of methods of measuring vapourdensity, there is not at present a convenient method when the tension is small. Hofmann's method requires for small tensions a cumbrous apparatus, whilst in Habermann-Dumas' method there is the objection of the additional determination of the weight of the vapour.

A tubular glass vessel of 100 c.c. capacity closed at one end is carefully cleaned and dried. Inside the tube is placed a short barometer tube. The substance is weighed in a small glass sphere having two sides drawn out into fine tubes; one capillary tube is stopped with alloy, and the other hermetically sealed after the introduction of the substance. The end of the glass vessel is drawn out into a capillary tube and the air is exhausted from the vessel; when the levels of the arms of the barometer tube remain unaltered the vessel is hermetically sealed. It is next immersed in a bath of known temperature, and the difference of level of the arms of the barometer is read off by a millimetre scale, so that the pressure is known. The vessel is immersed in water and the capillary point broken off. The vessel when filled is weighed. Thus, having measured the volume of a given length of the barometer tube, we have sufficient data to determine the vapour-density.

In order that the barometer may contain no moisture, it is constructed as follows:-A tube is bent into the form of a double U-tube, and one end is passed through a cork inserted in the neck of a fractional distillation flask which contains a small quantity of mercury. The flask is inclined, so that the end does not touch the mercury, and heated while dry air is drawn through the tube. When the flask is placed upright, the tube is immediately filled with mercury.

The proper quantity of substance to be placed in the sphere is determined by a preliminary calculation. For higher temperatures, the barometer is filled with an alloy of 3 parts lead and 1 part tin, which does not stick to the tube and has no perceptible tension.

For higher temperatures, the author also shows that a modification of L. Meyer's apparatus (Ann. Phys. Chem. [2], 1880, 550) can be employed.

C. S.

Viscosity of Dilute Aqueous Solutions. By S. ARRHENIUS (Zeit. physikal. Chem., 1, 285-298).-The solution is contained in a glass sphere of volume 0.9846 c.c., and is allowed to flow vertically through a fine capillary tube, the bore of which suddenly enlarges at its lower end, so that the velocity of efflux is negligible. The flow of liquid may be started by turning on a tap connected with the upper part of the sphere. The sphere and the capillary tube connected to it are immersed in a water-bath so that the solution may be kept at a known temperature. Since the liquid escapes with a very small velocity, the work done by the solution in falling through the height of the capillary tube will be entirely spent in overcoming the resistance due to viscosity, provided no energy is consumed in setting up vortex-motion as the liquid passes from the capillary tube into its enlarged end. In any case, the energy so spent will be less than if all the kinetic energy

of the liquid passing into the enlargement were employed in setting up vortex-motion. But on this hypothesis the energy spent may be calculated by a formula given by Hagenbach (Ann. Phys. Chem., 109, 385), and was found to be negligible with the dimensions of the apparatus used by the author. Thus a correction, which has been calculated by various experimenters on hypotheses never completely satisfied, is entirely avoided. By placing the capillary tube vertical, the small particles which are seen to form are carried away and not left adhering to the sides of the tube. Hence the coefficient of relative viscosity, 7, will be given by 7 = (st)/(ST), when s is the specific gravity of the solution, t the time of efflux of the volume contained in the sphere, and S, T similar quantities for water.

The results of numerous experiments show that the relative viscosity of a dilute aqueous solution of two different substances is equal to A* By, where x, y are the volume percentages of the two substances, and A and B two functions of the temperature which remain constant for the same temperature and the same substance. There does not appear to be any connection between A and the viscosity of the corresponding substance. The constant A diminishes as the temperature increases in the case of non-conductors, so that the values of the relative viscosity tend to a common limit, unity. The viscosity of water is increased by the addition of a small quantity of a non-conductor. Those normal solutions which have greater conductivity have in general less relative viscosity, but there does not seem to be any simple law connecting conductivity and viscosity.

C. S.

Dissociation of Hydrated Salts. By P. C. F. FROWEIN (Zeit. physikal. Chem., 1, 5—14 and 362-364).-The relation which must exist between the heat of combination of the water in hydrated salts and the maximum vapour-tension has been thermodynamically calculated, but hitherto has not been experimentally established in a satisfactory manuer. This is the aim, therefore, the author has in view.

The thermodynamical relation given by Van't Hoff, d. 1K/dT = q/21, is transformed, in order to be applicable to calorimetric work, into d. F/dT = Q/212, where T is the absolute temperature, Q the amount of heat evolved by the absorption of 18 kilos. OH, by the dehydrated salt, and F the ratio of the maximum tensions of salt and water. By integration of the above on the assumption that Q remains constant between small limits of temperature

This equation does not, however, give numbers agreeing with Thomsen's observed values for the heat of combination in the cases of CuSO, + 50H2; BaCl2 + 20H2; SrCl2+ 60H2; MgSO, + 70H2, and ZnSO, + 70H2, if the values for FF2 are calculated from the numbers given by Pareau and Wiedemann for the vapour-tensions.

Attributing this to errors in the vapour-tension determinations, arising partly from the very small differences in pressure which have to be observed, the author has redetermined these by means of an

apparatus in which olive oil is used in the manometer in place of mercury. The numbers thus obtained for FF, give values of Q which agree well with those observed by Thomsen for all the beforementioned salts, with the exception of SrCl2 + 60H2. The tension of Na2HPO + 120H2 has also been redetermined, and the calculated value for Q brought into agreement with the observed, a result which the former determinations by Debray and Müller-Erzbach had not rendered possible. H. C.

Nature of Chemical Affinity. By W. OSTWALD (Zeit. physikal. Chem., 1, 61-62).-In a former paper (Abstr., 1886, 294), the author has pointed out that chemical affinity depends not only on the nature and the relative distance of the atoms from one another, but also on the direction in which it acts. The difference in the molecular conductivities of the two nitrosalicylic acids, [COOH OH: NO2 = 1:23 and 1: 2: 5], is here offered as proof that although in each the relative positions of the NO, and OH to the COOH-group is the same, yet the former acid is the stronger of the two, the closer grouping of the radicles being more favourable to their combined action.

H. C.

Thermodynamical Expression of the Influence of Temperature on the Rate of Chemical Change. By F. URECH (Ber., 21, 56). A continuation of the author's investigations (Abstr., 1887, 768). Instead of Van't Hoff's expression, log c=-A/T + BT + C, the equation log c-A/T + B', corresponding with the thermodynamical formula log e-q/RT+ B', has been applied, and is found to give results agreeing generally with the author's observed values for the rates of inversion of saccharose with hydrochloric acid of different strengths.

c

H. C.

Action of Sulphurous Acid on Periodic Acid and Rate of the Change. By F. SELMONS (Ber., 21, 230—241).-The action of sulphurous acid on periodic acid takes place for all proportions of the two acids, the products varying, however, with the amounts employed. A separation of iodine takes place only within the limits represented by the equations

4H SO, HIO1 = 4H2SO, + HI

4H2SO3 + 4HIO1 = 4H2SO、 + 4HIOз,

namely, when more than one and less than 4 mols. HIO, enter into reaction with 4 mols. H.SO3, the production of iodine being in fact due to the action of the hydriodic on the iodic acid.

With dilute solutions, the rate of change of the reactions involving the production of iodine from sulphurous and periodic acids can be measured, using starch solution as an indicator. This reaction is similar to that studied by Landolt (Abstr., 1886, 658) on the action of sulphurous acid on iodic acid, and depends on the concentration of the solutions when the proportion between the two acids is kept constant, on the molecular weights for unit concentration, and on the temperature. If C, is the concentration of the sulphurous, Cp that of