Compressibility of Rock Salt. By F. BRAUN (Ann. Phys. Chem. [2], 33, 239-240).-The author in a recent note (this vol., p. 214), whilst giving the preference to Röntgen and Schneider's value 5x10 for the cubic compressibility of rock salt over the value 1.4 × 10 found by him, as his own determination was a comparatively rough one, pointed out that according to Voigt's measurements of elasticity the value would be 1·6 × 10. The author has since seen a later paper by Voigt (Ann. Phys. Chem. [2], 15, 497) in which he corrects his former result, as in obtaining it he made use of a torsion formula which he has since found to be incorrect. Voigt's later measurements would give the value 42 × 10, which, considering the difficulty of the piezometer method, agrees fairly with that obtained by Röntgen and Schneider. G. W. T. Rate of Transformation of Metaphosphoric Acid. By P. SABATIER (Compt. rend., 106, 63-66).-The conversion of metaphosphoric acid into orthophosphoric acid involves the introduction of an acid function of medium activity and an acid function of feeble activity, in addition to the original energetic function. Only the latter affects methyl-orange; the second feebler function affects phenolphthalein, the third can only be qualitatively recognized by means of the blue C4B. The acidity of the liquid h to methyl-orange remains constant, whilst its acidity to phenolphthaleïn increases, and affords a measure of the progress of the transformation. When conversion into orthophosphoric acid is complete = 2h, and at any stage of the change 2h gives the amount of metaphosphoric acid y still remaining unaltered. The metaphosphoric acid was obtained by dissolving phosphoric anhydride gradually in cold water, and also by strongly heating the ortho-acid, and dissolving the residue in water. Both preparations behaved in the same manner. These solutions were heated at different temperatures for definite intervals of time. At a given temperature, log y is proportional to the time. The velocity of the change at each instant is proportional to the mass of transformable substance present in the system, and hence log y x log a log b, and therefore y = ba, where b is the value of y at the commencement, and a is a constant which is a function of the temperature and the concentration, and increases with both, but the exact law of increase has still to be determined. C. H. B. Point of Transition and Point of Fusion. By J. H. VAN'T HOFF (Rec. Trav. Chem., 6, 36-42; 91-94; 137-139).-The particular reaction considered is the conversion of mixtures of sodium and magnesium sulphates into astrakanite, Na,Mg(SO,), + 4H2O, at temperatures below 21.6°. It is found that in order to separate sodium sulphate from the mother-liquors from sea-water, which contain a large proportion of magnesium chloride, it is necessary to cool the liquid below 5°, and hence it would appear that the system MgCl2+ Na,SO, is stable below 5°, but at higher temperatures the reciprocal system is formed. If a solution containing sodium sulphate and magnesium chloride in equivalent proportions is placed in a dilatometer, no anomalous expansion is observed at 5°, but if 2 mols. of sodium sulphate are present for each molecule of the magnesium salt, there is a very considerable expansion if the liquid is heated above 5°, and a considerable contraction if it is cooled below this temperature. This particular temperature is in fact the point of transition between the two systems, but the author has previously observed that magnesium chloride cannot exist in presence of sodium sulphate at temperatures below 5, because the tension of the water of crystallisation of the latter salt is less than the vapour-tension of the saturated solution of the two salts. The slow evaporation at a low temperature of a solution containing sodium chloride and sodium and magnesium sulphates in equivalent proportions yields a mixture of the three salts, but if evaporation takes place at a higher temperature cubical crystals of sodium chloride are obtained, together with rosettes of astrakanite, which is formed by the union of the two sulphates with elimination of water. In presence of sodium chloride, formation of astrakanite takes place at 6, and the reverse change at lower temperatures, whilst in absence of sodium chloride astrakanite is formed at 21-6° (Abstr., 1886, 968). The reduction of the point of transition from 216° to 5° by the presence of sodium chloride, corresponds with the reduction which this salt produces in the freezing point of its solution. The double acetate, CuCa(C2H ̧O2)、 + 8H2O, decomposes at 75° into the simple salts Ca(C2H3O2)2 + H2O and Cu(C2H3O2)2 + H2O, with elimination of 6 mols. H2O, and the change is accompanied by a distinct contraction; at lower temperatures, the double salt is formed, and expansion takes place. The volume changes are analogous to those which accompany the solidification of water, and it was to be expected that the effect of pressure would be analogous in the two cases. At the author's request, Spring has subjected the double acetate to high pressures at varying temperatures, and finds that under a pressure of 6000 atmospheres decomposition takes place at 40°. Under the same pressure at 16°, there is no appreciable decomposition if the duration of the experiment is short, but if pressure is prolonged, evidence of decomposition is readily recognisable. Under a pressure of 2000 atmospheres, decomposition takes place distinctly at 50° (compare this vol., p. 341). Van't Hoff points out that determinations of the minimum pressure required to produce reactions of this kind, together with measurements of the resulting changes in volume, will make it possible to express in kilogram-metres the absorption of energy involved in the reaction, and conversely the liberation of energy resulting from the reverse change. C. H. B. The Position of Atoms in Space. By J. WISLICENUS (Ber., 21, 581-585).- Lössen has challenged Van't Hoff and Wislicenus to state their views on units of affinity from the standpoint of their geometric theory, and he affirms that the question of the position of the units of affinity in space must be answered not after, but before, 2 e VOL. LIV. the question of the positions of atoms in space, and above all that a definition of " unit of affinity" must be given (this vol., p. 218). The author admits that his views quite bar the conclusion that atoms are material points, and it is impossible, therefore, to consider them as configurations in space in which the several chemical units of action of polyvalent elements can be localised. But this conclusion is admissible when the so-called atoms are considered, not as atoms in the strict sense of the word, but as simple groups of primitive atoms, similar to the compound radicles but less complex; and therefore the position of the elementary atoms in the molecule must be determined before Lössen's demand can be taken into earnest consideration. The author has shown that the determination of the arrangement in space of the atoms in a molecule is within the domain of experi ment, and he points out that the study of the configuration of the molecule is the only way to solve successfully the question of the distribution in space of its spheres of action-the so-called units of affinity. Only by studying the properties of molecules can inductive conclusions be drawn as to the properties of atoms. The author's views are stated as follows:-It is more probable that the atoms are configurations in space composed of atoms of primitive elements than that they are simply points of energy; the most likely assumption is, therefore, that the atoms are comparable with the compound radicles, and, as in the latter, their units of affinity are located in certain parts from which they exercise their action. It is possible, in time, to succeed in obtaining definite ideas not only of the form of elementary atoms, but also of the relative posi tions of their spheres of action. It is not impossible that the atom of carbon resembles more or less in form a regular tetrahedron, and that the causes of those actions, which are evidenced by the units of affinity, are concentrated in the angles of this tetrahedric configuration, perhaps for analogous reasons, and similarly to the electric action of a charged metallic tetrahedron. The real carriers of energy would finally be the primitive atoms, exactly as the chemical energy of compound radicles is the resultant of the energy of the elementary atoms. F. S. K. The Atomic Weights of the Elements. By A. BAZAROFF (J. Russ. Chem. Soc., 1887, 61-73).-The author finds that the variation in the numbers expressing the atomic weights of the elements, arranged according to the periodic system, is analogous to the changes in the properties of the elements and their compounds. When the most probable numbers for the atomic weights of the elements are arranged according to the periodic law, and, either in the horizontal or in the vertical series, the atomic weight of an element is divided by the atomic weight of the element with the next lower atomic weight, products are obtained which decrease regularly with increase in the atomic weights compared. In the horizontal series, the maximum is Be 9.08 found to correspond with the relation of = = 1.2953, the Li 7.01 minimum to 207-5 = Bi 10054. This decrease, however, is not continuous, the products alternately decreasing and increasing, as is seen from the following examples : : This relation is best represented graphically, the order of atomic weights standing as abscissæ, and the aforesaid quotients as ordinates: in this way a curve of zigzag form is obtained. This relation is expressed by the author in the form of a law, namely, that "the increase in the atomic weights of the elements proceeds with a variable intensity, the smaller coefficient of change varying with the larger in such a way that both regularly decrease." Another regularity is observed in the vertical groups, for example, with the coefficients in the second group : -- This change the author expresses by the law: "In the vertical series in the periodic system, the relation of two neighbouring atomic weights decreases with increasing atomic weight, but this decrease is alternately larger and smaller." Neither of these functions extends to the whole periodic system, as there are many elements whose atomic weights are not yet determined with sufficient accuracy. Yet it is possible, from the data which are to hand, to state the following general law: "The magnitude of the atomic weight of each element is determined by the magnitude of the atomic weights of the elements next to it in the periodic system both horizontally and vertically." The author further analyses the exceptions to the said rules, and expresses the opinion that several of them are due to the atomic weights not being determined with sufficient accuracy, but it is difficult to say how far he is right in correcting several of the better determined numbers (he assumes for TI 202 instead of 2037!). Notwithstanding the apparently complicated character of the relation pointed out, the author thinks it possible that when the fundamental data are more exactly determined, it may be possible to calculate the atomic weight of an element with greater accuracy than is the case at present. B. B. Raoult's Method of Determining Molecular Weights.* By V. MEYER (Ber., 21, 536-539). In investigating some derivatives of benzil, two series of isomeric compounds were obtained; both series have the same constitution, in the ordinary sense of the word, and are yet distinct from one another and yield different derivatives. The * Raoult's papers on this method are Abstr., 1883, 7, 278, 952; 1884, 254, 701, 808, 952, 1248; 1885, 122, 858; 1886, 197, 763. Compare also this vol., p. 361. Ph C R isomerism of the two series is expressed by the formulæ Ph CR Ph CR R:C Ph. and By means of a modification of Raoult's method for determining molecular weight by measuring the amount by which the solidifying point of a solvent is lowered by a known weight of a dissolved substance, it is shown that the molecular weight of the compounds of the two series in question is identical. The method employed, which only involves the use of an ordinary thermometer divided into 0·1° and a simple apparatus, will be described. N. H. M. Application in Chemical Laboratories of Raoult's Method for Determining Molecular Weights. By K. AUWERS (Ber., 21, 701-719). The use of Raoult's method for determining molecular weights by V. Meyer (preceding Abstract) led to a consideration of the conditions necessary for its successful application. The exception which many solutions, notably those of inorganic salts in water, show to the law put forward by Raoult (Abstr., 1886, 763), renders this law useless for practical purposes, and the relation M = T/A, where M is the molecular weight, T the molecular depression, and A the depression caused by 1 gram of substance in 100 grams of solution, is used. It is recommended to first determine the value of T, which will be constant for the series of compounds under consideration, by taking members of the series of known molecular weight. A measurement of A then suffices to determine M in subsequent cases. A great objection is the impossibility of working with relatively small amounts of material, for although the above relation does not hold for concentrated solutions, the solutions must not be too dilute, and some discrimination is necessary in adjusting the strength to give the depression which should be measured in each case. With regard to the solvents used, it is of course a condition that no chemical action should take place between the one selected and the substance under examination. The selection of a suitable solvent is a matter of great importance. Water is generally objectionable on account of its tendency to form hydrates and its non-solvent action on most organic compounds. As it is necessary also, according to Raoult, that the depression measured should not be less than 0.5°, as the relation does not hold when it is allowed to sink below that value, a relatively large quantity of material is necessary for each determination. Benzene gives better results, but cannot be used in the case of alcohols, phenols, or acids, and introduces an element of uncertainty in dealing with substances allied to these. By far the best solvent, and the one that can be used with the greatest range of substances, is glacial acetic acid. The errors in this case are small, and the measurement of a depression of about 0.3° suffices. A great advantage also is that this solvent admits of working at ordinary temperatures. It is, however, found necessary, owing to its |
