1891.] the Discharge of Ley den Jars. 37 



Here the maximum is not quite so well marked, and it is a little 

 uncertain on which side of the 15-cm. position it lies. 



Take u = 15 however, and add the capacity of the leads, 147, to 

 that of the jar, 357, and the calculated wave-length comes out 47 

 metres ; or, without the capacity of the leads, 39| metres. 



This is hopelessly different from 76-J metres, but it is probable 

 that it is the upper octave. To make the fundamental wave-length 

 79 metres would require u to be about 130 cm. 



It is evidently desirable in these experiments that the jar capacity 

 shall be much greater than the leads capacity; for when the two 

 capacities are comparable in size, the recoil kick, although it still 

 exists, has no very well denned maximum. 



It would also be desirable in future experiments on the same plan 

 to use, as dielectric in the condenser, a substance, like paraffin, 

 whose specific inductive capacity may be depended on as fairly 

 constant. 



Experiments with Sliding Condenser. 



41. The following experiments, made with an adjustable con- 

 denser, were intended to examine the effect of varying the capacity 

 as well as the self-induction ; also to see what evidence of harmonics 

 could be detected. 



The condensers used were each of them a pair of silvered tubes 

 sliding into one another ; their capacity, as the inner tubes were 

 pushed into six different graduations, being measured above ( 34). 



The discharge circuit was a pair of No. 17 B.W.G. copper wires 

 each 2 metres long, stretched horizonally on glass pillars, parallel to 

 one another and 10 cm. apart. One of the tube condensers was 

 arranged as a bridge at one end of these wires ; and the discharger, 

 with its knobs set 1 cm. apart, was used as a movable bridge and set 

 at different measured distances from the condenser. 



Charging from the machine was done through wooden sticks, so 

 as not to introduce unknown capacities. The "longer leads " lead 

 from the coats of the tube-condenser to the spark-micrometer, where 

 the equivalent A spark and the maximum C or recoil spark were 

 observed. 



The numbers entered in the following table are the excess of the 

 max. C spark over the virtual A spark, i.e., C spark-length minus A 

 spark-length, in millimetres. The maximum occurs in the different 

 columns at about u = 50, say 35, 30, 25, 20, and 15, respectively. 



