The peroxide penetrates deeper and deeper into the plates as they are successively used as positives, and by repeated reversals their surfaces ultimately come to have a porous spongy condition - a state which takes a month or two to attain. Returning to a consideration of the failure of the negative plate : on taking a piece of amalgamated lead and withdrawing it upon exhaustion, the mercurized surface of the plate is seen to be dimmed over by a glaze. This glaze, it can be further shown, is sulphur, and the early failure of the negative plate is due to the formation of a coat of sulphate of lead. The film is extremely thin, imperceptible on ordinary lead, yet is sufficient to stop the current. That this is the correct explanation is shown by wiping off the film by rubbing the plate on a cloth, when the current again passes till. the sulphate has been redeposited. At first sight it would seem best to employ a metal for the plates, of which the sulphate was soluble; but that would not answer, for the positive plate, when the current was reversed for the peroxide, would also be dissolved, and the peroxide deposited on the positive plate conducts electricity, whereas the sulphate of the negative plate does not.

The improvement introduced by Faure, was to get reducible plates without the great number of preliminary reversals necessary under Plante's system. Plante provided oxidizable plates, but their oxidizability did not extend to any great depth, and most of the force escaped. Faure decided to coat his plates with some porous peroxide, and found that minium, red - lead, and litharge were suitable for the purpose. By this means a large absorbent surface was provided, and the waste occasioned by the giving off of gas from the cell was delayed, thus increasing the storage capacity of the cell by that interval. The peroxide coat sponged out till all was peroxidized, and so there was no discharge till the plate was full. For the positive plate it was found best to use minium, and for the negative red - lead as a coat. If the quantities of coats and liquid were properly adjusted, both plates would be filled together; but, as a matter of fact, this was not attained. That was the Faure principle, in which there were no seversals of current, and the economy in time over the Plante or reversing system could be estimated from the fact that a Faure cell was prepared in a week, whereas the Plante cell took 2 or 3 months to fill, while there was no essential difference in the results of the 2 forms of storage.

The time and expense of production were thus greatly diminished under the Faure system. But in practice it was not found easy in the Faure cells to make the coating adhere to the plates for any length of time. The red - lead or minium had no affinity for the lead steeped in the liquid, and speedily peeled off, and the difficulty arose, how could the 2 substances be kept together. The first attempt at a solution was to place porous material between the plates, and then the composition was literally tied on to the plates by bands of cloth. But this did not give a good chemical contact with the plate, and another difficulty was that the cloth was rapidly attacked and decomposed by the dilute sulphuric acid. To prolong its life, the best trousering cloth was used, and this proved very expensive, and did not last long nor give a good contact; besides this, the coats and plates would not bear such shaking as they would sustain in transit by railway.

A further improvement was devised, and consisted in putting the oxidizing coating into perforations or interstices of the metallic plates themselves. This was hit upon by Swan and Sellon almost simultaneously, Swan being actually the first in point of time. In the perfected battery known as the Faure - Sel - lon - Volckmar, the lead becomes a grid filled with an oxidizable composition. First of all, cast - iron moulds are prepared, scored all over like a gridiron with a network of straight channels. Two of these matrices are put together, and the molten lead is poured into them; rectangular pierced sheets of lead are cast as the result. These framed grids are then filled with composition - litharge for the positive and red - lead for the negative plate, the composition being made into a pulp with diluted sulphuric acid, and run in. Though the plates may be somewhat dusty, the composition adheres well to them, and can only be removed by bending the plates back. The next stage is to form cells. A large number of plates are put together in a trough, the positives and negatives being kept apart, and each plate is separated from the next by a frame of india - rubber. The dilute sulphuric acid is then poured in, and the plates are slowly peroxidized.

They gradually become blackened upon the lines of the lead - work, the centres of the holes being the last portions to change colour, and the whole surfaces are in a spongy condition, accessible to the liquid. A peroxidized plate has a soft crystalline appearance of a deep brown colour. The plates ore then stored in Faure cells ready for use, being carefully isolated by the india - rubber bands. Another plan is now being tried by Thornton, of pushing out the composition from certain holes, and filling them up with projecting india - rubber plugs; but if the plugs are not bent, no contact is possible under the present plan. The state of saturation in the liquid is ascertained by an ingenious copper balance and hygrometer. The plates are made in 2 sizes, one double of the other, and are packed together, 9 positives and 9 negatives, in a box. The small sizes of cell are known as 1/2 h. - p., and the large ones as 1 h. - p. cells, the term signifying, of course, such a power "per hour," and combinations of cells are manufactured up to 5 h. - p., the only difference being in the number of plates per cell. About 16 amperes per hour can be stored in a small pair of plates, and such a pair will discharge a current for 6 to 8 hours, giving forth 160 amperes.