236 PROCEEDINGS OF THK AMERICAN ACADEMY. 



the oscillations in the secondary, at least when there is no spark gap 

 in the secondary circuit. 



The sides of the rectangle (Figure 5) were reduced to a few centi- 

 meters in length, so that it could be safely assumed that its period was 

 much shorter than that of the vibrator. The plates of the vibrator 

 were fixed at 61 cm. apart, and the side KM of the rectangular 

 secondary was placed at 6 cm. from the conductor joining the plates 

 of the vibrator, with its centre opposite the spark gap. The sides 

 K L and M N lay in the horizontal plane through the axis of the 

 vibrator, and were held by the end supports at 30 cm. apart. The ap- 

 paratus was symmetrical about a vertical plane through its spark gap 

 normal to the axis. The exploring terminals were kept at L N and 

 bolometer readings taken for each small addition to the length of the 

 sides K L and M N. When best resonance was obtained with the 

 shortest length of the secondary circuit that gave a maximum, it was 

 assumed that the secondary had the same period as the primary, and 

 that its equivalent length was a half wave length, its actual length 

 depending upon the effect due to the free ends. The occurrence of 

 resonance is a very marked phenomenon, even with a vibrator that 

 damps as rapidly as the Hertzian. The following table shows two 

 series of readings for the first maximum when an iron wire was 

 used : — 



Length of side of rectangle 15 

 Deflections ...... 107 



Deflections 94 



There can be no free motion of the electricity at the ends of the 

 secondary circuit, but an accumulation alternately positive and negative 

 and the resulting alternation of potential, the phase at L being always 

 opposite to the phase at N. Elsewhere along the circuit the electri- 

 city moves with more freedom and less accumulation. The point O 

 may be called the electrical middle of the circuit, where the accumula- 

 tion is least and the movement most restrained. The electromotive 

 impulses from the vibrator act directly upon the side K M, so that O 

 remains a point of free motion, or the central segment of the wave, 

 while L and N are always places of no electric motion, or the nodal 

 points. The nodes under this view are the places of greatest poten- 

 tial difference, so that in the graphic representation of tlie results the 

 maximum points of the curves correspond to the nodes, the bolometer 

 throws being the largest when the exploring terminals are placed 

 where the potential difference is greatest. The shortest circuit being 



