stant (which is the character of a combination) as soon as there are not more than 600 volumes of the gas-a quantity which corresponds to equivalent of hydrogen for 1 equivalent of palladium. From this moment the hydrogenized palladium behaves as a definite combination, susceptible of dissociation, and of which the tension does not depend on the temperature-after the manner of carbonate of lime, studied by M. Debray.

In order to exhibit more distinctly this invariability of pressure which is characteristic of combination, we will cite the results of an experiment made at the temperatures in the vicinity of 160°.

Series of observations made between 20° and 180° have taught us that within these limits of temperature the phenomenon retains the same characters; that is to say, the tension remains constant as soon as the quantity of fixed hydrogen is no more than 600 volumes, which corresponds to the formula Pa2 H. We have thus obtained the following numbers, which permit us to construct the curve of the dissociation tensions of the compound:

*We have therefore been able, in the experiment cited, to take away two thirds of the gas without diminishing the tension. When the last traces of Pa H have been decomposed, the palladium remains saturated with dissolved hydrogen gas. We then reenter the conditions of a phenomenon of solution; and fresh subtractions of gas should bring a rapid decrease of the pressure: this is evidenced by the three following determinations::

Remarks.-The numbers in this Table were obtained with a palladium wire; this was preferred to a sheet for these determinations, because the boundary tension is more rapidly reached. Besides, a sheet leads, although more slowly, to the same results. The cast and the forged metal present no difference in respect of the tensions; and the behaviour is the same, whether the metal has been charged by the pile or by a prolonged stay in hydrogen.

The above numbers show that Pa2 H does not sensibly emit gas at the ordinary temperature. The hydrogen-tensions which limit the decomposition of the compound increase at first slowly; but from 140° very small additions to the temperature are attended by a rapid augmentation of the pressure. Between 130° and 140° the tension becomes equal to the pressure of the atmosphere.

This compound, then, cannot be prepared at a higher temperature than 130°, unless the hydrogen be compressed. The most convenient temperature for its production is that of boiling water, at which the dissociation tension is below one third of the atmospheric pressure.

In brief, our experiments prove that palladium forms with hydrogen a definite compound, of which the formula is Pa2 H. This combination, once formed, can dissolve hydrogen gas after the manner of platinum, and in quantity variable with its physical state. This property of Pa H explains the difference between the numerical results obtained by Graham according as he employed palladium wire or spongy palladium.

In another communication we will show that potassium and sodium form combinations with hydrogen which have the formulæ K2 H, Na2 H (K=39, Na=23). These, together with hydrogenized palladium, form a series parallel to that of which M. Wurtz discovered the first term in the combination of copper with hydrogen Cu2 H2 (Cu=63.5), and to which he gave the name of hydride of copper.-Comptes Rendus de l'Acad. des Sciences, vol. lxxviii. pp. 686-690.

THE

L. On the new Contact Theory of the Galvanic Cell.
By J. A. FLEMING, B.Sc., F.C.S.*

HE contest that has for so long been waged between the supporters of the two theories of the galvanic cell, the contact and the chemical, can hardly be said to have been brought even now to a decisive issue. For although the contact theory, as originally proposed by Volta, received a fatal blow when the law of conservation of energy became clearly understood, yet in its place a new contact theory has arisen, supported by novel and important experimental evidence, which has again been placed by recent writers on electrical science in formidable opposition to its old rival.

The old contact theory of Volta had its origin in an entire ignorance of the science of energy. It simply referred the current produced through the circuit of a pile to the effect of the metallic contacts, and it ignored the thermal and chemical changes which are also necessarily present; but it had to be finally abandoned when once it became clearly understood that the appearance of a current involved the disappearance of some other energy, actual or potential, as an invariable accompaniment. The new contact theory may be said to have had its source in the discovery of Sir W. Thomson, that there is undoubtedly a difference of potential produced when dissimilar metals are placed in contact-a fact not only abundantly proved by Thomson by direct experimental evidence, but, as he has

* Read before the Physical Society, March 21, 1874. Communicated by the Society. 2 D

Phil. Mag. S. 4. Vol. 47. No. 314. June 1874.