8 



Chant — Experimental Investigation into the 



tuted for them, while the connections remained precisely the 

 same. 



The spark gap was always the same for the cylinders, being 

 about 0-T ffim long. 



The spherical oscillators were of the Righi pattern. The 

 spheres C, D (fig. 3), were held in circular discs of hard rub- 

 ber, in which openings had been turned to receive them, by 

 melting paraffine about them. The smaller knobs, K, K, were 

 on the ends of jointed rods, R, R, which were held in hard 

 rubber rods ~N, N, rigidly fastened to the disks carrying the 

 spheres. By this means the knobs could be adjusted to the 

 spheres and would remain so when the spheres were adjusted 

 to each other. To allow this latter to be done one disc was 

 held to the horizontal support beneath by a screw passing up 



JUL^. 



through the support into it, while the other could be slid 

 along the support by a screw. The entire arrangement is 

 shown in fig. 4. The spheres were of three sizes. With those 

 2*5 cms and 4 cms in diameter, knobs K, K of diameter 19* m were 

 used. When the spheres had a diameter of 10 cms , the knobs 

 had a diameter of 37 mm . 



The membrane, m, m, was first softened by soaking for a few 

 minutes in water and glycerine, and was then securely bound 

 to the circumference of the hard rubber discs by cords which 

 rested in grooves made to receive them. The membrane was 

 such as is used as a container for the familiar " Bologna sau- 

 sage," and was obtained at a large packing house. It made a 

 perfectly oil-tight vessel, fig. 4, the largest oscillator being in 

 action for hours without a single drop of oil escaping. 



The oil used was described as white liquid vaseline. 



When in operation the oscillator was placed in the parabolic 

 reflector on a suitable support, seen in fig. 5, and was con- 

 nected to the coil by wires, L, L, passing through tubes as in 

 case of the cylinders. 



