Figure l. — Skid and beam portion of the beam trawl. The towing bridle is attached to the upper leading edge of each skid (points B). 

 The headline is attached to points N. The lateral angle iron extensions (S) are intended to act as stops to prevent wheel turning when off 

 bottom. The skids are lined with netting (hatched areas). The main net is not shown, but the footline is represented by F. 



Mass.) time-depth recorder was attached to the 

 trawl, but was lost after only 17 trials. 



Results 

 Consistency of the Two Wheel Counts 



Wheel count ratios ranged over a wide span of 

 values. However, there was good agreement be- 



tween the wheels on about 50*^ of the hauls (Fig- 

 ure 2) with a ratio of 1.105 or less, and SS'^^ of the 

 ratios were <2.000. Even though the histogram of 

 ratios showed that close agreement between 

 wheels was more common than poor agreement, it 

 was difficult to determine the normal range of 

 random variation. A high ratio might have been 

 due to inherent variation or to mechanical failure 

 of one of the wheels. A wheel might have recorded 



40, 



30. 



20 



10- 



50 PERCENTILE 



■A Wi^y^ PyTxW myrr\ f 



56 RATIOS > 2.00 



.30 1.40 1.50 



HIGH WHEEL COUNT 



Y"^ ""^ , r-^'l Fl Fp_JLa — filfi^rrrri — ,i in f\ 



r 



I 90 



1.00 



.60 



.70 



1.80 



2.00 



LOW WHEELCOUNT 

 Figure 2. — Frequency of occurrence of highrlow wheel count ratios, n = 337. 



793 



