No doubt could exist as to the best source for the production of such a strong current as that required, the difficulty was to select the dynamo machine which would be most suitable for the purpose. The machines manufactured by Siemens and Halske of Berlin had been very highly spoken of as yielding good results in the precipitation of copper on a large scale at Oker in Germany. This firm was appealed to, and one of their small machines was recommended, which, although calculated to perform only about one-half the work required, was deemed sufficient to commence with. This machine is numbered C16, and it is stated to give a current of 240 amperes with an electromotive force of 15 volts, and an expenditure of 7 horse-power. (Of course this is considerably higher than the calculated power required.)

From the given capacity of the machine it is easy to calculate that the best result would be obtained, without danger of over-heating the machine, by arranging the total resistance to be equal to 1/16 ohm. As the subject of the resistance of various liquids to the passage of the electric current has been so imperfectly investigated, and data respecting other factors being likewise unobtainable, it was impossible to determine the best form, number, and arrangement of baths for the purpose in question. The best plan was evidently to make the resistance of each bath as little as possible, consistent with facility of working, and then to arrange them parallel, or in series, or both, so as to include the resistance mentioned above. Eight baths were decided upon. These were made of wood, lined with rubber, and each had the following internal dimensions. Length 150 cm., breadth 70 cm., depth 100 cm. The thickness of the wood was 5 cm., that of the rubber 3-4 mm. Each four baths were formed by the division of one large tank, 3 m. by 1 1/2 m., and the two tanks were firmly secured by iron bolts running from side to side. Placed longitudinally they occupied exactly the breadth of the building, and they were fixed at the end of the same, at an elevation of one metre from the ground.

In front was a platform, at one side of which was a door for the admission of the cuttings, after being washed and packed. The dynamo was situated in the engine room just behind the baths, and communication was made by two copper cables passing through a hole in the wall. At either side wall, and on a level with the baths, was a dissolving tank capable of accommodating half the cuttings after the removal of the tin. A little farther on, and nearly on a level with the ground, were the evaporating down tanks. The crystallising tanks were situated beneath the ground level, so that the solution could be run off from one stage to another without any pumping arrangement.

The anodes were composed of the tin scrap. Baskets were obtained to pack the cuttings in. These, at first, were made of wicker-work, but as they were too flexible, and soon rotted by the action of the acid, their place was supplied by strong wooden baskets, whose sides were formed of stout upright wooden bars, 2 cm. thick, with spaces between them sufficient to allow the solution to circulate freely, while preventing the exit of the scrap. The internal dimensions of the baskets were: - Length, 120 cm.; breadth, 30 cm.; depth, 85 cm. Great care was required in packing the scrap, because, if it were packed too closely, the metallic surfaces thereby united, and preventing the circulation of the electrolyte between them, necessarily retained their coating of tin. These baskets held 60-70 kilo, of the scrap, the 8 being capable of accommodating about half the total quantity required. Long and narrow strips of the same material were employed to complete the communication with the conductor. As the resistance of iron to the current is comparatively great, a large number of these were required to prevent excessive heating.

At one end they were soldered together, and connected with the copper conductor by means of binding screws; the other extremities were distributed throughout the scrap.

Copper plates were employed as cathodes. These had a thickness of 1 1/2 mm., were 120 cm. long, and 95 cm. in breadth. There were 16 in all, two for each bath, placed one on either side of the baskets. To keep these thin plates of copper as plane as possible, each was surrounded by a framework of copper rod of square section. They were coated with tin to prevent corrosion, as well as to avoid solution by any accidental reversal of the current. These plates rested in grooves at the sides of the tanks, placed at a distance of 10 cm. from the sides of the baskets. They were provided, as also were the baskets, with rubber rollers extending to the sides of the baths, enabling them to be raised out of the same with ease, and without injury to the rubber coating.

Dilute sulphuric acid formed the electrolyte. This was employed, not only on account of its comparatively small resistance to the current, but also because it was convenient to turn the solution into the iron sulphate tanks, as soon as it became saturated with that salt, and all the tin had been precipitated from it. Commercial acid of 60° B. was diluted with 9 volumes of water.

Above the tanks was a pulley arrangement for raising the baskets and plates out of the baths as required; there was also an arrangement of levers and eccentrics constructed, whereby the baskets were kept in gentle motion in the baths, thus exciting circulation in the liquid, and tending to prevent polarisation. The horizontal axis upon which the eccentrics were disposed made about two revolutions per minute, thereby raising the baskets a distance of about 5 cm. Levers were fulcrumed into the wall; these passed over the eccentrics, and at their extremities ropes were fixed communicating with the baskets.

The current was conducted by thick copper wires of several plies. Although already coated, they were enclosed in rubber tubing as an additional protection.