Page images

more nearly saturated with gas, and on connecting them both show a transient H-polarisation, which is soon changed to O-polarisation.

The polarisation of either electrode increases as its surface is diminished; that is, it is proportional to the density with which the gas is separated.

An investigation of the rate of loss of both H- and O-polarisation, on open circuit, showed that this is never uniform, but reaches maximum and minimum values. Two explanations of this are suggested. One is, that the union of the occluded gases with the metal is not a union in fixed proportions; and hence the dissociation may show maximum and minimum values. Another rests on the following considerations:-The motion of the occluded gases in the substance of an electrode must always take place in the direction of least saturation, and during charging must be from without inwards. The moment the circuit is broken, a rapid escape of gas takes place from the surface; and there will then be at some little depth a layer of metal highly charged with gas, having on each side of it less saturated layers. For some time afterwards, therefore, the motion will take place in two directions. A series of wave-like motions of the gas will then ensue, which accounts for the maxima and minima. The shorter the duration of the charging current, the more rapidly the gas condensation diminishes from without inwards, and the sooner the rate at which polarisation disappears reaches its maximum. A comparison of these rates for the anode and cathode leads the author to the conclusion that the union of platinum with oxygen is of a looser nature than that with hydrogen, and the motion of the former gas within the electrode more free.

Nevertheless, even when the polarising current has been of brief duration, and the cathode not previously O-polarised, the loss of H-polarisation still reaches a maximum after a distinct interval. This points either to the normal presence of oxygen in the platinum (possibly absorbed during the preliminary heating to redness); or, as Streintz suggests (Ann. Phys. Chem. [2], 13, 644; and 17, 841), to pure platinum being electronegative towards the same metal containing hydrogen.

The results with gold electrodes are essentially the same as, although somewhat more complex than, those with platinum. They point to the occlusion of gases by this metal for two strengths of battery electromotive force, namely, between 0-2 and 0.7 and above 1·1 Daniell. The presence of air in the voltameter diminishes the force required for occlusion.

The results with palladium are explained by the enormous occlusion of hydrogen, which takes place principally for forces between 0:4 and 1·4 of a Daniell, and by the superficial oxidation of the metal. But even here occlusion of oxygen probably takes place. Only after long passage of a strong current (one chromic acid element) does palladium become completely saturated with hydrogen; and the discharge of this gas when the circuit is broken may then last for weeks.

Warming the voltameter may diminish O-polarisation by scattering the gas; in the same way, it may diminish H-polarisation when this is high, but may increase it by assisting the escape of occluded gas when the polarisation is feeble. Сн. В.

Action of Ethylene Bromide on Alkyl-metallic Oxides: Preparation of Acetylene. By DE FORCRAND (Compt. rend., 104, 696-699).—The author's researches on the action of ethylene bromide on alkyl-metallic oxides dissolved in the corresponding alcohols show that the reaction C2H,Br, liquid + 2ROM dissolved in nROH = 2MBr precipitated in n + 2ROH + C2H2 gas, develops in the cases of the first five alkyl potassium oxides +15.92; +19.30; +2244; +32.94; +27.32 Cal. respectively, and in the cases of the first three sodium compounds +136; +886; +13.66 respectively. The corresponding reaction with the solid compounds develops +4146; +46-48; +45 94; +50-72, and +44-50 Cal. respectively in the case of the potassium alkyl oxides, and +3174; +29·18; +33·2; +34.92; +35 02 Cal. respectively in the case of the sodium compounds.

In all these reactions monobromethylene is also produced. The heat of formation of this compound is not known, but the difference between it and y, the heat developed in the above reactions, is greater the greater the value of y. It follows that at the same temperature the proportion of acetylene in the product will be greater, and that of monobromethylene less, the greater the value of y, a conclusion which is confirmed by experiment.

The greatest development of heat is observed in the case of potassium isobutyl oxide, and this compound can be used for the preparation of acetylene. Potassium is dissolved in isobutyl alcohol, the excess of alcohol distilled off in a current of dry hydrogen, and the residue heated at 200° in hydrogen, and after cooling, mixed gradually with ethylene bromide. The reaction takes place at the ordinary temperature, and the gas is purified from isobutyl alcohol and monobromethylene by passing it through absolute alcohol; 500 to 600 c.c. can be obtained from 2 granis of potassium. Potassium ethoxide can also be used, but the gas contains a greater proportion of monobromethylene.

In 1861, Sawitsch (Compt. rend., 52, 157) obtained acetylene by the action of ethylene bromide on sodium amyl oxide in sealed tubes at 100°; and from potassium ethoxide and propylene dibromide he obtained allylene. C. H. B.

Heat of Formation of Tartar Emetic. By GUNTZ (Compt. rend., 104, 699-707).-The heat of formation of potassium hydrogen tartrate was determined by dissolving equivalent quantities of normal potassium tartrate and tartaric acid in dilute hydrochloric acid.

CHO solid + CHOK2 solid = 2C,H,O,K

[merged small][merged small][merged small][merged small][ocr errors]

The heat developed by the action of tartaric acid on the normal tartrate is of the same order as that developed by the action of sulphuric acid on the normal sulphate (+75 Cal.).

The heat of formation of tartar emetic was determined by dissolving equivalent quantities of potassium hydrogen tartrate and antimony oxide in dilute hydrofluoric acid.

*CH1O10,KO,HO solid + SbO, solid = HO solid+CHO10, KO,SbO, solid.......

develops -0.85 Cal.

The heat of solution of the anhydrous salt at 12° is -51; of the hydrated salt -53, whence

*C,H,O10,KO, SbO, solid + HO solid = CH,O10,KO,SbO,,HO solid....

develops -0.5 Cal.

In the process of drying, tartar emetic should not be heated above 100°, since at 110-120° it loses more than one equivalent of water. When heated at 180°, it loses two equivalents of water, as Dumas observed, and forms an anhydride which dissolves in water with development of heat and production of a solution identical with an ordinary solution of tartar emetic.

*C,H2O, KO,SbO, solid + H2O2 solid = C.H,O10, KO,SbO, solid

develops +41 Cal.

C. H. B.

Influence of Double- and Ring linking on Molecular Volumes. By A. HORSTMANN (Ber., 20, 766-781).-In this paper, the author has brought together and classified in numerous tables the values obtained by various investigators for the molecular volumes (both at 0° and at the boiling point) of such substances as seemed likely to throw light on the influence of molecular structure on the molecular volumes of organic compounds.

Comparing compounds with open carbon-chains, the author finds that when a double linking is converted into a single linking by the addition of two atoms of hydrogen, the molecular volume increases, the increase at 0° varying from 41 to 94, the mean increase being 64. The instances taken include the conversion of unsaturated into saturated compounds, as well as of highly unsaturated compounds into less unsaturated (such as diallyl into hexylene). Most of the differences observed lie closely around the mean, 6'4.

Next, taking substances containing the benzene-ring and comparing their molecular volumes with those of their hexahydrides, the mean difference is found to be for 6H = 224, or for H2: = 7.5, or nearly the same as in the case of the open-chain formation. But when by a further addition of H, we pass from the hydrides of the aromatic compounds to fully saturated compounds the increase is much greater, the mean for H, being 16.2. This latter passage is of course accompanied with the change from the ring to the open formation. Similar results are shown between hydrides of aromatic compounds and their isomerides with open chains. In this case-where there is only difference of constitution and not of amount of hydrogen presentthe molecular volumes of the compounds with open chains exceed those of the isomeric benzene compounds by about 8.5. Similar results are obtained by a comparison of compounds of the naphthalene series (containing two rings) with corresponding compounds of the aromatic and fatty series.

* These equations are given as in the original paper, O = 8.

[ocr errors]

Turning to the thiophen series, no such easy comparisons can be made since no hydrides are known. But comparing thiophen with diethyl sulphide, we find the molecular volume of the latter (1076) to be 30-4 in excess of that of thiophen (772). Now here we have an addition of 3H2, a change of two double linkings to single ones, and of the ring to the open formation. Taking the numbers derived from the benzene series, we should obtain (7·5 × 2) + 16·5 31.5, or almost exactly the difference found. Thus, although as we do not know the intermediate hydride we cannot speak with certainty, the numbers seem again in the thiophen series to point to a greater difference being caused by the addition of the two hydrogen-atoms which cause a breaking up of the ring formation, than is caused by a like addition when only a change from double to single linking takes place.

With regard to compounds containing nitrogen, we find in the fatty series that the conversion of nitriles to amines by the addition of 2H2, gives a mean increase of 118 or 2 × 5.9. In the case of pyridine-derivatives, comparison is more difficult, since the possible variation of mode of linking of the nitrogen-atom comes into consideration. But comparing the pyridine compounds with their hexahydrides we find the increase 3H2 = 168 or 3 × 56, and further comparing these hydrides with the corresponding amines of the fatty series we find the mean increase, H2 = 18.6.

The above numbers all refer to molecular volumes determined at 0°. If those determined at the boiling point are employed, similar results are obtained, although of course the actual numbers are slightly different. In this case, the mean values are as follows:-I. Where the change is simply from double to single linking, H2 in the fatty series = 71, in the aromatic series = 82, in the naphthalene series 80, in the nitrogenous fatty series 36, in the pyridine series 65. II. Where the change is from the closed ring to the open chain H, = 20.2.




From the above results, the author draws the conclusion that "unsaturated compounds with ring formulæ have much smaller molecular volumes than their isomerides with open chains and polylinkings." He considers that in the present state of our knowledge of the subject, and absence of information as to the disturbing influences of other constitutional variations, no great stress should be laid on the actual numerical amounts of these differences, and that for the same reason it is also not well at present to attempt to

calculate absolute values for the elements. The results obtained with compounds of the aromatic, thiophen, and pyridine series seem to speak in favour of Kekulé's ring formula.

L. T. T.

Effect of Sulphuric Acid on the Solubility of Sulphates. By R. ENGEL (Compt. rend., 104, 506-508).—When sulphuric acid is added to solutions of sulphates which do not form acid salts, the solubility of the sulphate is diminished, but not in the same way as the solubility of chlorides in hydrochloric acid; that is to say, one equivalent of the acid does not precipitate one equivalent of the salt. The sulphuric acid behaves as if each molecule combined with

12 mols. of water and prevented them from exerting any solvent action. When the quantity of acid is very large, the divergences from this law become more marked. Details are given in the case of copper and cadmium sulphates. Other sulphates behave in a similar manner. The solubility of zinc sulphate is diminished by the first 12 equivalents of sulphuric acid, and it is only beyond this point that an acid salt begins to be formed. Magnesium sulphate and some others also show this phenomenon.

C. H. B.

A Particular Case of Solution. By F. PARMENTIER (Compt. rend., 104, 686-688).-Crystallised phosphomolybdic acid containing 234 per cent. of water dissolves in ether with development of heat, but if excess of ether is then added, and the liquid agitated, the ether and the ethereal solution of the acid will not mix, but separate into two distinct layers. If the crystals were not dry, or if the ether was not really anhydrous, water separates as a third intermediate layer. When the ether is evaporated, the phosphomolybdic acid crystallises in the ordinary form. If the ethereal solution is separated from excess of ether, and heated in a sealed tube, a quantity of ether separates as a supernatant layer, the volume of which increases as the temperature rises. After cooling, the supernatant layer can again be mixed with the rest of the liquid, and a homogeneous fluid is


An ethereal solution of phosphomolybdic acid saturated at 13° has a sp. gr. of 13, and is soluble in alcohol in all proportions, but will not dissolve in water. If an aqueous solution of phosphomolybdic acid is agitated with ether, the whole of the acid is removed from the water, and the ethereal solution separates and sinks to the bottom. If the crystals of the acid are allowed to effloresce, or are heated at 100° for some time, they afterwards only dissolve in ether if a sufficient quantity of water is added.

The solubility of the acid in ether increases with the temperature, the quantity dissolved by 100 parts of ether being as follows:

[blocks in formation]

The solution of 1 kilo. of the acid in ether develops 22.8 Cal. or 90-2 Cal. per gram-molecule. A solution saturated with ether can dissolve large quantities of the acid with development of heat, but with separation of water as a supernatant layer.

C. H. B.

Solubility of Calcium Orthobutyrate and Isobutyrate. By J. CHANCEL and F. PARMENTIER (Compt. rend., 104, 474-478).Chatelier (Abstr., 1885, 340, 473) has deduced a relation between the heat of dissolution of a substance and its solubility, according to which, if dissolution takes place with absorption of heat, the solubility increases with a rise of temperature, and vice versa. The authors have determined the solubility of normal calcium butyrate, which diminishes with a rise of temperature, and of calcium isobutyrate, which increases with a rise of temperature below 80°.

Calcium orthobutyrate forms nacreous lamellæ of the composition

« PreviousContinue »