Of DENISON tTNlVERSITY. 1^ 



wave-length for wires 8 cm apart, and 162.1; for wires 3 cm apart, a 

 change of 9 cm, or more than 5 per cent. 



It became desirable next to study the effect of surrounding a por- 

 tion of the parallel wires by a containing vessel such as would be suita- 

 ble for holding a liquid. I used a covered trough of zinc, 100 x 10 

 X 10 cm, with the wires passing centrally through rubber stoppers in 

 the ends, one end being made to coincide with the centre maximum. 



The internodal spaces, which had been equal before, were now 

 different, that which included the metal box being shortened 3.3 cm or 

 about 2 per cent. 



The result is of the sort that we might expect, as the proximity of 

 the metal box would naturally increase the capacity of a given length 

 of the wires in that neighborhood, rendering a shorter length necessary. 



Later experiments, in which there were two internodal spaces be- 

 fore the box was reached, developed the fact that the influence of the 

 box upon the half-wave external, but adjacent to it, was considerable, 

 since this half-wave was invariably 5 cm shorter than the one remote 

 from the box. 



When the box was filled with distilled water only a small, constant 

 deflection of the galvanometer was obtained, and this was shown to be 

 due partly to a direct magnetic effect of the induction coil upon the 

 galvanometer and partly to current induced in the bolometer wires 

 from the wires connecting the induction coil with the storage battery. 



Up to this time the little " Leyden jars " had always been placed 

 at the remote end of the parallel wires. The jars were now removed, 

 and still smaller ones of the same sort placed at the ventral segments 

 of the two external half waves. Although these had very small outer 

 coatings — consisting of cylinders of copper foil 5 mm in diameter and 

 5 mm long — it was found that their capacity could not be neglected, 

 but was equal to 2,5 cm wire-length. This appeared from the fact that 

 the internodal space containing the "Leyden jar" was shortened 2.5 

 cm, the other remaining the same as before. 



Still smaller jars were next made of a single turn of very fine wire 

 about a glass tube 5 mm in diameter. These were found to make no 

 appreciable change in the position of the maxima, and still, when placed 

 at a ventral segment and connected to the dynamo-bolometer, suffic- 

 iently large deflections (100 mm) were obtained. 



Very fine resonance systems were obtained in the external portion 

 of the parallel wires in those experiments where a node was forced at 



