312

INVENTIONS OF NIKOLA TESLA,

more. At a certain point a minute increase of the pressure willcause the glass to crack. The loss of energy involved in splittingthe glass may be practically nothing, for though the force is great,the displacement need be but extremely small- How imaginethat the glass would possess the property of closing again per-fectly the crack upon a minute diminution of the pressure.This is the way the dielectric in the discharge space shouldbehave. But inasmuch as there would be always some loss in thegap, the medium, which should be continuous, should exchangethrough the gap at a rapid rate. In the preceding example, theglass being perfectly closed, it would mean that the dielectric inthe discharge space possesses a great insulating power; the glassbeing cracked, it would signify that the medium in the space isa good conductor. The dielectric should vary enormously inresistance by minute variations of the e. m. f. across thedischarge space. This condition is attained, but in an extremelyimperfect manner, by warming the air space to a certaincritical temperature, dependent on the e. m. f. across the gap,or by otherwise impairing the insulating power of the air. Butas a matter of fact the air does never break down disruptiveiy ,if this term be rigorously interpreted, for before the suddenrush of the current occurs, there is always a weak currentpreceding it, which rises first gradually and then with compara-tive suddenness. That is the reason why the rate of change isvery much greater when glass, for instance, is broken through,than when the break takes place through an air space of equiva-lent dielectric strength. As a medium for the discharge space, asolid, or even a liquid, would be preferable therefor. It is some-what difficult to conceive of a solid body which would possess theproperty of closing instantly after it has been cracked. But aliquid, especially under great pressure, behaves practically like asolid, while it possesses the property of closing the crack. Henceit was thought that a liquid insulator might be more suitable as adielectric than air. Following out this idea, a number of differentforms of dischargers in which a variety of such insulators, some-times under great pressure, were employed, have been experi-mented upon. It is thought sufficient to dwell in a few wordsupon one of the forms experimented upon. One of these dis-chargers is illustrated in Figs. 168a and 1685.

A hollow metal pulley p (Fig. 168a), was fastened upon an ar-bor a, which by suitable means was rotated at a considerable