This is the cistern as opposed to the constant supply system, and it is probable that, as in water services, the latter, by which a central store supplies the force, will be most popular. The central supply of electromotive force will be equally available for incandescent, arc, and other forms of lamp at the same time. The excellence of the result is due to the remarkably low internal electrical resistance of the cells, and in this respect they differ from any batteries. In these cells the internal resistance is as small as •0016 of an ohm, rising to .002 of an ohm when nearly exhausted. Where a gas - engine is employed it is not important, as with a steam - engine, to keep it continuously at work, and with this motor the secondary battery should be used as a regulator rather than a store. A waterfall is an instance of a motor where economy demands that it shall be employed continuously; by the use of secondary batteries, a 70 h. - p. fall would provide a 100 h. - p. light at night. For domestic lighting, the problem is how to reduce the electromotive force so as to render it perfectly safe for household purposes. Dynamos have a high internal resistance: secondary batteries will give the same amount of electromotive force with far less resistance.

For An Energy Of 2000 H.p., the resistance would, with the cells, be less than 2 ohms, but would rise to about 400 ohms with a dynamo. A high electromotive force is most economical; but to obtain this directly from a dynamo, the rate of revolution must be enormous, and the friction is correspondingly great.

On the formation and construction of lead batteries, some valuable notes have been contributed by J. T. Sprague to the English Mechanic, mainly as follows.

Plante's directions for charging are to effect 6 or 8 reversals of current the first day, prolonging the successive charges; this is continued next day till the duration of useful charging becomes a couple of hours; at this limit it becomes necessary to give intervals of rest between the charges, during which a local action takes place: these rests require gradual prolongation to several days, and even weeks; after the cell approaches the capacity of storage intended, the current should not be reversed, but the cell should simply be charged and discharged as if in actual work. The process of this forming, therefore, occupies some months, except with very thin plates, and if pushed too far, the whole substance of the plate may be converted into peroxide, which would render it liable to break up, and would increase its resistance.

Heat assists the formation and reduces the time required for the process; therefore during the process of charging the temperature may be advantageously raised to 100° to 160° F. (38° to 71 °C), and allowed to fall as soon as the charge is effected. But heat is objectionable in actual working, because it facilitates the oxygen and hydrogen assuming the gaseous state and going off to waste.

Alcohol added to the extent of 5 Per cent. to the acid solution is said to assist the formation. Berliner states that it requires but an hour to develop a heavy oxide surface capable of taking a large charge.

Nitric acid is recommended by Plante, who says that by soaking the plates for some hours in nitric acid mixed with an equal volume of water, he has greatly reduced the time of formation. The effect is to produce a porous surface more quickly acted upon; but it - is evident that for this treatment the plates should be thicker.

Electro - deposited lead has been tried for the same object. But lead is a very troublesome metal to deposit; unlike other metals, it does not spread as an even film, but will dart out in fine arrows from points on the surface, which either fall off as they lengthen, or close the circuit to the other plate; for this reason the presence of lead salts in solution, or the use of acids, etc, which will dissolve lead, is very objectionable.

Amalgamation of the lead has been employed by Paget Higgs. It is by no means clear that it is an advantage on the whole : it must tend to weaken the lead, and therefore to fracture of the plates. It is not desirable or useful at the peroxide plate, because it resists the formation of Pb02, and tends to form oxide or salts of mercury; the mercury also prevents the molecular union of the lead and peroxide, which latter therefore tends to separate from the plate. It is probable that mercury facilitates the absorption of hydrogen at the leaden plate; but even there the advantage is very doubtful, because in regular working the real action which goes on is the reduction of the lead sulphate which has been formed during discharge.

The plates may be either flat plates interleaved like those of a condenser, or they may be large plates rolled up as cylinders, or folded up together. Flat plates give a simpler construction, and have the great advantage that they can be " formed" in a separate vessel and combined as desired; each can also be removed singly in case of injury, and easily replaced. They have a serious disadvantage, however. It is evident that the molecular volume of lead salts exceeds that of lead, so that there is constant expansion and contraction going on, which tend to produce bulging surfaces, and this much more readily upon flat parallel plates. It is therefore desirable to introduce slips of glass or other insulating material to resist this, and prevent the plates coming in contact. The cylindrical form is made by laying a long sheet of lead on a table, placing upon it a number of strips of soft vulcanized rubber arranged diagonally upon the lead, then another sheet of lead with similar diagonal strips; the sheets are then rolled up firmly so as to form 2 parallel spirals separated from each other. The insulating strips can be inserted as the rolling up proceeds.

The thickness of the lead must be sufficient to bear its own weight and the strains put upon it after use; the peroxide plate should therefore be about double as thick as the other. In large batteries, lead 1 millimetre ('03937 in.) thick is used, or about 24 sheets to the inch; 1 sq. ft. of lead 1 in. thick weighs 59.1 lb., therefore this would be about 2 1/4 lb. lead. The smaller cells are made of lead little thicker than that prepared for damp walls, which is about i to 5 oz. per ft. For ordinary purposes, it is probable that sheets of 1 1/2 lb. and 1 lb. would be most advantageous, with the formation carried so far as to convert 1/2 lb. of the lead to peroxide on the one plate.