104 Miss H. Brooks on the Damping of the 



in a high state of polish, as the spark shows a disinclination 

 to pass from any point on the ball which has become rough- 

 ened or tarnished. It prefers to leap a longer distance from 

 a smooth place. 



To investigate the effect of increase of capacity on the 

 damping, several jars were connected in parallel and placed 

 in the circuit. The damping was found to be practically the 

 same for one, two, four, eight, or twelve jars. The relation 

 for smaller capacities than that of the leyden-jar employed 

 in the first experiments, *. e., less than 2500 E.S. units, was 

 obtained by means of an ebonite condenser whose capacity 

 could be varied as desired by removing or adding to the 

 tinfoil coating. It was found that the damping changed 

 very slightly for capacities over 1000 C.G.S. units, but below 

 that point the damping increased rapidly with decrease of 

 capacity (see fig. 6). 



This points to the same conclusion as that indicated by the 

 curves for spark-length and damping with different capacities, 

 that the damping reaches a comparatively steady condition 

 when the current is large. 



Change of pressure at the spark-gap had a very noticeable 

 effect on the damping. The terminals of the air-break were 

 enclosed in an air-tight glass bulb, which was connected to 

 an air-pump. The air was thoroughly dried before entering 

 the bulb. 



The damping for any given spark-length decreases as the 

 pressure becomes less. An exactly similar effect was ob- 

 served when hydrogen and carbonic acid were introduced 

 into the bulb instead of air. For a given spark-length and 

 pressure the damping in hydrogen is less than in air. In 

 fig. 7 is given a curve for pressure and damping with a spark 

 of 1 cm. in air, and also for a spark of 3 cms. in hydrogen. 



Observations point to the existence of a critical pressure 

 depending on the spark-length and the nature of the gas, 

 below which the damping begins again to increase, evidently 

 with great rapidity, for the resistance soon becomes so great 

 that no discharge at all is obtainable. This point has not 

 been very closely investigated, as the arrangement employed 

 was not sensitive enough to be sufficiently affected by these 

 very small currents. With a more sensitive arrangement 

 accurate measurements might easily be made. 



It has been shown that a very small part of the damping 

 is due to the resistance of the leads, and in a leyden-jar 

 circuit the loss of energy from radiation is also very small, so 

 that by far the greater part of the dissipation of energy must 

 take place at the spark-gap. 



