22 C. A. Chant — Variation of Potential, etc. 



present, but it is not nearly so intense as in either of the other 

 two systems. 



4. The most effective length of the antenna, therefore, is 

 one quarter-wave-length, not a higher multiple. 



5. When inductance is inserted between the condenser cir- 

 cuit and the earth the fundamental oscillation is not so regular 

 or intense, other oscillations (overtones) being superposed. 



6. For the production of oscillations by the direct method a 

 small capacity cannot satisfactorily balance the antenna ; in the 

 inductive method, however, a capacity acts like an earth con- 

 nection or a similar wire. 



IV. Continuation of Former Investigation. 



In the previous experiments with Hertzian plate oscillators 

 of various sizes and with wires ranging in length from 300 to 

 860 cms , there was usually one "chief" minimum of potential- 

 variation between 100 and 200 cms from the free end, and 

 always a marked one about 10 or 15 cms from the other end of 

 the wire. It was hoped that by employing longer wires the 

 phenomena of standing waves Would be much better exhibited, 

 and that several "chief" minima would be shown. Such, 

 howevef, has not proved to be the case. Wires 2050 and 

 4090 cms long were carefully explored, the action on the wire 

 being produced by means of an oscillator having plates 40 cms 

 square and the straight connection between 60 cms long, but the 

 only unmistakable minimum was approximately 150 cms from 

 the free end, the same as was perfectly formed with wires 

 from 300 cms upwards. 



Some evidence was obtained as to the cause of the marked 

 minimum near the other end of the wire. It is due to the 

 direct action of the oscillator on the detector. As described 

 in the other paper, an attempt had been made to allow for this 

 direct action by taking the reading when the wire was in place 

 and also when it was removed, and then subtracting the latter 

 from the former. This assumes that the two effects are quite 

 independent, but such seems hardly to be the case. In the 

 former experiments the detector lay in a horizontal plane on 

 the top of a carriage which was moved along the wire. Thus 

 the detector's length was parallel to that of the oscillator, 

 though the little wing was perpendicular to it. As described 

 in Part II of the present paper, the detector was now hung in 

 a vertical plane from the wire, so that its length was perpen- 

 dicular to the axis of the oscillator. With this arrangement 

 the minimum disappeared, thus showing that it had been pro- 

 duced by the direct action of the oscillator on the detector. 



University of Toronto, Toronto, Canada. 



