Let us insulate, now, the outer coating of a Lane's unit jar _L_, the jar just described, and put in connexion with it the inner coating of a jar _F_ exteriorly connected with the earth (Fig. 30). Every time that _L_ is charged with +_q_, a like quantity +_q_ is collected on the inner coating of _F_, and the spontaneous discharge of the jar _L_, which is now again empty, takes place. The number of the discharges of the jar _L_ furnishes us, thus, with a measure of the quantity collected in the jar _F_, and if after 1, 2, 3, ... spontaneous discharges of _L_ the jar _F_ is discharged, it is evident that the charge of _F_ has been proportionately augmented. [Illustration: Fig. 31.] Let us supply now, to effect the spontaneous discharge, the jar _F_ with knobs of the same size and at the same distance apart as those of the jar _L_ (Fig. 31). If we find, then, that five discharges of the unit jar take place before one spontaneous discharge of the jar _F_ occurs, plainly the jar _F_, for equal distances between the knobs of the two jars, equal striking distances, is able to hold five times the quantity of electricity that _L_ can, that is, has five times the _capacity_ of _L_.[29] [Illustration: Fig. 32.] We will now replace the unit jar _L_, with which we measure electricity, so to speak, _into_ the jar _F_, by a Franklin's pane, consisting of two parallel flat metal plates (Fig. 32), separated only by air. If here, for example, thirty spontaneous discharges of the pane are sufficient to fill the jar, ten discharges will be found sufficient if the air-space between the two plates be filled with a cake of sulphur. Hence, the capacity of a Franklin's pane of sulphur is about three times greater than that of one of the same shape and size made of air, or, as it is the custom to say, the specific inductive capacity of sulphur (that of air being taken as the unit) is about 3.[30] We are here arrived at a very simple fact, which clearly shows us the significance of the number called _dielectric constant_, or _specific inductive capacity_, the knowledge of which is so important for the theory of submarine cables. Let us consider a jar _A_, which is charged with a certain quantity of electricity. We can discharge the jar directly. But we can also discharge the jar _A_ (Fig. 33) partly into a jar _B_, by connecting the two outer coatings with each other. In this operation a portion of the quantity of electricity passes, accompanied by sparks, into the jar _B_, and we now find both jars charged. [Illustration: Fig. 33.] [Illustration: Fig. 34.]