among them does not lead to the separation of hydrogen. In order that there may be no polarisation, in other words, that the electromotive force may be the same whatever the strength of the current, the materials of the cell should be such that the passage of the current and the accompanying chemical action cause no change in the nature of the substances in contact. In proportion as polarisation is avoided, a battery or cell is called constant. No perfectly constant form of battery has yet been devised, but there are two or three forms which make a fair approach to constancy when properly constructed and employed. Daniell's battery (1836) was the earliest form of constant battery, and is still one of the most efficient. The metals employed are

amalgamated zinc and copper, the zinc dipping in a dilute solution of zinc sulphate or in dilute sulphuric acid, and the copper in a saturated solution of copper sulphate. The two liquids are kept from mixing by a porous partition of unglazed earthenware, or sometimes by putting the heavier saturated copper solution in the bottom of a vessel and the lighter liquid above. A cell arranged in the former way is represented in Fig. 99, where c and z are the copper and zinc plates respectively, and D is the porous partition. During the passage of the current, metallic zinc is converted into zinc sulphate, which dissolves in the liquid in contact with the zinc plate, and metallic copper separates from the solution of copper sulphate and is deposited on the copper plate. The metallic surfaces thus retain permanently their original

character, but the concentration of the solutions changes, especially in close contact with the metal plates, and this causes a small decrease of electromotive force while the current is passing. The normal electromotive force of a Daniell's cell is about 1.07 volt.

In Grove's battery (1840) the evolution of hydrogen is suppressed by the employment of strong nitric acid, which gives up part of its oxygen and converts the hydrogen into water. One of the metals is again amalgamated zinc dipping into dilute sulphuric acid (1 vol. of acid to 8 or 10 of water), which is separated from the nitric acid by a partition of porous earthenware. The use of nitric acid necessitates the replacement of copper by a metal, platinum, which the acid does not attack.

Bunsen's battery (Fig. 100) is a modification of Grove's, in which rods of dense gas-graphite are substituted for the expensive platinum plates. All else remains as in Grove's battery.

The electromotive force of a Grove's or Bunsen's cell is about 1.8 volt.

Among the many other forms of cells that are in more or less common use, we may mention the bichromate and the Leclanché cells. In both of these the plates consist of amalgamated zinc and graphite respectively in the former, the plates dip in a solution of potassium bichromate acidulated with sulphuric acid; in the latter, the liquid is a saturated solution of ammonium chloride (sal ammoniac), and the carbon plate is closely packed

round with black oxide of manganese in small fragments. The electromotive force of a bichromate cell is about 2 volts, but it has the disadvantage that the liquid acts on the zinc even when no current is passing; it must not, therefore, be left to stand with the plates in the liquid. A Leclanché cell has an electromotive force of not quite 1.5 volt on open circuit. It falls considerably

when a current passes, but recovers when left to itself, and the plates may be left in contact with the liquid for long periods without injury.

In all the cells mentioned the zinc terminal is negative.

Fig. 101 represents the mode of connecting a number of cells to form a battery, and shows a conducting wire joined to opposite terminals of the first and last cells.

CHAPTER XI.

ELECTRIC CURRENTS-RESISTANCE-OHM'S LAW.

108. The Electric Current.-We have already said (§ 103) that when the terminals of a battery are connected with the plates of a condenser, positive electricity passes to the plate in connection with the positive terminal, and negative electricity to that connected with the negative terminal. This accumulation of electricity on the two plates continues until each of them is brought to the same potential as the pole of the battery with which it is connected. When this state of things has been reached, there is a condition of equilibrium in which one condenser-plate, the corresponding wire, and the battery terminal to which it is attached are all at one potential, and the second condenser-plate, wire, and terminal are all at another potential, differing from that of the first plate by an amount equal to the electromotive force of the battery, —that is, the same change of potential occurs between the two sides of the dielectric of the condenser as between the two terminals of the battery, and the tendency of the condenser to discharge exactly balances the electromotive force of the battery.

If, however, the terminals are connected by a continuous wire, a condition of statical equilibrium is never established. The wire tends to bring the two terminals to the same potential, while the battery tends to maintain between them a difference of potentials equal to its own electromotive force. These two tendencies cannot be satisfied simultaneously: the result is that an intermediate state of things is reached, in which there is a continuous passage of positive electricity in one direction through the wire, and of negative in the opposite direction, tending to equalise the potential at the two ends; while, at the same time, the battery continues to supply positive electricity to, and to remove negative from, that end of the wire which has the higher potential, and to supply negative and remove positive at the other end, thus keeping up a difference of potentials between the ends. This difference of

potentials falls short of equality with the electromotive force of the battery to an extent to be investigated later (§ 109).

As long as these conditions are maintained, the wire possesses certain special properties which are briefly indicated by saying that it is traversed by an electric current. These properties are not in all respects the same in both directions—that is, some of them differ according as, when we pass along the wire in a given direction, we are proceeding from the positive terminal of the battery to the negative, or the reverse. The same thing is commonly expressed by saying that the properties of the wire depend on the direction of the current, which is conventionally understood to mean the direction in which positive electricity is conceived of as passing round the circuit, namely, through the battery from the negative terminal to the positive, and through the connecting wire from the positive terminal to the negative.

Experiment shows that the characteristic properties associated with what is called the passage of a current are possessed by every point in the circuit formed by the battery and the conductor by which the terminals are connected, and, further, that the intensity with which these properties are exhibited, as well as their direction, is the same all round the circuit. It is found also by experiment that the potential at any one point of the circuit remains constant, which shows that there is no accumulation of electricity at any part. Hence it is concluded that the current, of which the circuit is the seat, is of the nature of a continuous circulation such that every section of the circuit is traversed simultaneously by the same quantity of electricity. The quantity of electricity which passes any section of the circuit in a given time determines what is called the strength of the current. The unit of current-strength, or, as it is also called more shortly, the unit of current, commonly employed for practical purposes, is that which corresponds to a coulomb per second. Such a current is called a current of one ampere. The strength of a current expressed in amperes therefore gives the number of coulombs which pass every cross-section of the circuit in one second.

If the strength of a current is constant, it is represented by the formula

C = q/t

where q is the quantity of electricity which traverses any section of the circuit in t seconds, and since the quantity which traverses