14 BULLETIN OF THE LABORATORIES 



the box end, — better than had been obtained with the parallel wires 

 alone. This result is not attributed to the presence of the box, but 

 rather to the use of the Leyden jars at intermediate ventrals segments 

 instead of at the ends and to the readjustment of bridges already placed 

 whenever the capacity of the system was changed by adding a new 

 bridge. 



Thus in one experiment with bridge fixed at box end 255, nodes 

 were located as follows : 



-1 14.5 57 85 124 1.S5 1G9 190? 255 



Notice iiofies at 14.5, i:!.') and ■J.i.t show intervals 120 and 120. .t; iiipan 120 3. 

 " —2, 85, lOy, 255 " " 86, 84, 87; mean 85.7. 



" 57, 124 " •' ti7; 



Thus the 9 nodal positions fiill into three groups, showing intervals 

 of 12.3, 85.7 and 67 respectively. 



Now 120.3 X 2)^ = 301 



85.7 X 3>|' = 300 

 and 67 X 4)^ = 302 



/. e., according as the nodal position occupied belonged to group r, 2 or 

 3, the vibrating system consisted of 2i4, 31^ or 4^^ half waves, the 

 system of wires and end plates being ecjuivalent to 301 cm of straight 

 parallel wires. 



These results gave a simple means of calculating the parallel-wire 

 equivalent of the secondary plates and their connecting wires. As the 

 parallel wires began at —8 of the scale, the first node ( — 2) is 5 

 cm from one end, but the whole part of the system beyond this 



80 



node ^ — or 43 minus 5 = 2)^ as the equivalent in cm of the parallel 



2 



wires for the capacity of each plate (6x6 cm) and of the 1 1 cm of wire 

 leading from it to the parallel wires. 



From another system of nodes we get in the same way the same 

 result. Thus : 



120 



= 60 minus (7-(-i4 5) == 38.5 or practically the same as before. 



2 



From another experiment with a system of very different length of 

 parallel wires, the wire equivalent of the same plate came out 39.2 

 and 40 ; mean, 39.6. 



