by any other. These lines are determined at right angles to the resultant of the curve forces, indicated by the arrows over which the needle is placed their positions will consequently appear to vary with almost every variation in the length of the needle employed.

This curious distribution of magnetic polarity, or whatever force it may be that actuates the needle, by the present arrangement, is decidedly peculiar to the direction of motion indicated by the exterior arrow, Fig. 11; there being no similar distribution, as in Fig. 12, by simply reversing the direction of revolution. If, however, the magnetic poles and the direction of motion be both reversed at the same time, then there is, on the upper surface, a distribution of force in every respect similar to that represented by Fig. 11. Hence, if in Fig. 12 the north pole of the magnet were to be placed above the disc, instead of the south pole as there represented, the distribution of force would be indicated by the two systems of arrows in that figure; the revolution of the disc being in the direction of the large exterior arrow.

Now, as every condition, both of arrangement and motion, has been considered to be inverted to produce the distribution of force represented by Fig. 12, that figure may very well represent the lower side of the plate turned upwards, when the conditions of arrangement and motion are represented by Fig. 11. Indeed it is more convenient to examine the two sides of the disc in this manner; for when the needle is placed below, its motions cannot be very well observed, except at a short distance within the edge.

When the plate is not very large, this force is more equably distributed over the surface, as indicated by the distribution of the arrows in Fig. 13; but in no case is it exactly so.

I have examined the distribution of magnetic polarity in discs and 6ther forms of metallic surfaces with a good deal of attention, whilst the magnetic poles were variously posited with regard to them, and and I have collected a number of curious facts; many of which are exceedingly difficult to arrange, on account of the singular windings of the force which actuates the needle. I have, however, succeeded in tracing the distribution in some instances by experiments, which will be described in an early communication.

THE ANNALS

OF

ELECTRICITY, MAGNETISM,

AND

CHEMISTRY;

AND

GUARDIAN OF EXPERIMENTAL SCIENCE.

MAY, 1843.

Contributions to the Chemical History of Palladium and Platinum. By ROBERT KANE, M.D., M.R.I.A. Communicated by FRANCIS BAILY, Esq., V.P.R.S., &c. &c. &c.

(Continued from page 253).

Of the Nitrates of Palladium.

It has been long known that palladium dissolves in nitric acid without any evolution of nitric oxide gas unless heat be applied. On evaporating the olive-brown solution so obtained, it becomes more reddish, and if set aside when of a sirupy consistence, and allowed to cool slowly, the nitrate of palladium crystallizes in long rhombic needles, which are, however, so deliquescent that I found it impossible to determine with accuracy the quantity of water of crystallization which they contain. In general, the solution of this salt dries down to a mass with scarcely any trace of crystalline

structure.

If a solution of nitrate of palladium be dilluted with much water, a dark brown powder falls, which is a basic nitrate. It may be generated also by adding solution of potash, or water of ammonia in small quantity, to a solution of the metallic salt. This basic nitrate, when heated, evolves water, and then red fumes of nitrous acid; leaving oxide or suboxide of palladium, according to the temperature to which the material may have been finally exposed.

The following analyses will show that the basic nitrate, as prepared by different methods, is really of uniform constitution. A. Basic nitrate prepared by water of ammonia.

60 808 grains of this specimen were placed in a tube of Bohemian glass, about twelve inches long, and in front of this, but completely separated by some rolled pieces of platinum foil, the tube was filled for a space of about four inches, with clean finely-divided metallic Ann. of Elec., Vol. X, No. 59, May, 1843.

Χ

copper, as reduced from the oxide by hydrogen gas. To each end of the tube was attached a bulb-tube, containing recently fused chloride of calcium, and that next the metallic copper was placed, by a caoutchouc connector, in communication with a vessel of water, by the flowing out of which a current of air might be established through the apparatus, precisely as is effected in the process proposed by Liebig for drying organic substances previous to analysis. The long tube containing the palladium salt and the metallic copper being placed in a charcoal furnace, that portion containing the copper was heated to redness, and then, whilst by the flowing out of the water a stream of air was brought through the tube, a gentle heat was applied to the basic nitrate of palladium. Water and red fumes were given off, which latter were reduced to the state of nitrogen or nitrous oxide by contact with the ignited metallic copper. The current of air, which had been accurately dried by passing through the first chloride of calcium tube, carried these products forwards. The water was collected by the chloride of calcium tube into which it passed, whilst the gases mixing with the general current of the air passed into the vessel from which the water flowed.

As soon as the palladium salt had been feebly ignited, the process was interrupted, and the tube allowed to cool. It was then so cut by a file, as that the residual oxide of palladium could be removed without any sensible loss. It weighed 44.620 grains, or 73.15 per cont. The chloride of calcium tube in which the water had been collected, was weighed before and after the operation. The increase of weight was 7.208 grains, indicating 11.85 per cent. of water.

In order to verify the degree of oxidation and determine the quantity of metal in the residual oxide, the 44 620 grains were ignited in a current of hydrogen gas, and the water so formed was `collected in a chloride of calcium tube. It weighed 6·251 grains, or 10.28 per cent., containing 914 of oxygen, and the remaining metallic palladium weighed 38-927 grains, corresponding to 64.01 per cent.

B. Basic nitrate produced by water.

I. 29'665 grains of the brown precipitate, formed by diluting a strong solution of nitrate of palladium with a large quantity of water, gave, by very gentle ignition, a black residue of oxide weighing 21.747, or 73.31 per cent., and by vivid ignition 18.939, or 64:36 per cent. of pure metallic palladium.

II. 12.858 grains were mixed with copper filings, and more clean metallic copper being placed in front, the tube was heated in the same manner and with the same arrangement of apparatus as already described in A. The water collected in the chloride of calcium tube weighed 1420, corresponding to 11.20 per cent.

III. 17.239 grains were mixed with copper turnings, and placed at the bottom of a tube of Bohemia glass, which was then filled with clean freshly-reduced metallic copper. The tube was about