At the point of installation, the stream was 66 feet wide; 

 its average depth varied from 9-l/U to 21-l/U inches during the period 

 of experimentation with the device . The array of electrodes, which 

 was 96 feet in overall length was mounted diagonally across the stream; 

 a conventional weir-trap was installed between the upstream end of the 

 array and the stream bank (See Apple gate and Smith, 195l, for "portable 

 weir" trap design). The trap was shielded on its streamward side by an 

 electrically "grounded" metal screen. 



The electrodes in each row were connected in parallel in 

 order that the three rows might be energized in various combinations 

 (Fig. 7)" The "B" (center) row of electrodes in all cases was made the 

 common connection for a voltage of 110 VAC to the "A" (upstream) row and 

 a voltage of 55 VAC or 110 VAC to the "C" (downstream) row. Thus, oper- 

 ation as a uniform field barrier or as an incremental field barrier was 

 possible. In addition, the voltage gradient within either field could 

 be controlled to some extent by changes in the spacing of electrodes in 

 the individual rows. 



In certain of the experiments involving incremental fields, a 

 non-lethal "warning" field between the "B" and "C" rows of electrodes 

 was achieved by the application of 55 VAC, or one-half the AC line volt- 

 age, across these two rows of electrodes, A 60 cycle, 2 ; 1 stepdown 

 autotransformer was used for this purpose. In other similar experiments, 

 the effect of reduced voltage gradients in the downstream half of the 

 field was produced by increasing the spacing of "C" row electrodes to 

 8-foot centers and connecting this row to the same source of voltage as 

 the "A" or upstream row. 



The Kewaunee River electromechanical weir was tested both as a 

 two-row and as a three-row system over a wide range of field intensities 

 by the use of various combinations of electrode spacings and voltages. 

 Following is a tabulation of the combinations tested: 



iU 



