the total counts for the half -month period were greatest at 11 aom. No 

 reason for this variation could be found; it may have resulted from chance 

 since the run this year was the smallest of any year for the highest half- 

 month period at the Bradford ladder and few fish were counted each day. 



In three of the years, 1939, 19^1, and 19U2, the counts were 

 higher for the first hour of counting in the morning than for the second. 

 This probably resulted either from shad moving up during the night or from 

 shad that were already in the ladder the night before when counting ceased 

 at 8 Pom. 



In the same way, the percentage hourly counts for the Washington 

 ladder are plotted in figure 9. The years 19^3 and 19$0 were orrdtted because 

 the tobal counts were exceptionally small (table 3), The greatest irregu- 

 larity is in the curve for 19lih, No reason could be found for this anomaly. 



The pattern of counts in the Washington ladder is not quite as 

 uniform as that for the Bradford ladder. This probably occurs because the 

 coijits are smaller than those of the Bradford ladder. In general, the 

 counts were low during the morning hours, reached a peak in the afternoon, 

 and dropped off at night. In this resnect they are similar to those for the 

 Bradford ladder. 



Only during 1938 and I9I48 were fish counted at night at Bonneville. 

 In those years an extra shift was added that counted fish from 8 p.m. to 

 h a.m. During 1938 the counts were made from July 1 to September 15 in both 

 ladders but on alternate hours only. In I9U8 the night counts were mads for 

 each hour, but only during the first half of July and the first half of 

 August. 



Very few shad \-jere counted in 1938 during the night shift. At the 

 peak of the run (July I6 to 31) at the Bradford ladder a total of 6I8 shad 

 were counted during the l6-hour daytime period, vjhile only 8 shad were 

 counted during the h night hours. Counts at the Washington ladder during 

 the night were very small also. Strangely enough, the best night counts of 

 the season were made at both ladders in the first half of September when the 

 day counts were low. The total number of shad counted at night at the 

 Bradford ladder was 90 (It hours each night) while during the day the counts 

 for the same period totaled only 193 (counting done I6 hours each day). At 

 the Washington Shore ladder the night count for the same period totaled Iil 

 shad, while during the two day-shifts only 26 shad were counted. No explana- 

 tion for this anomaly has been found. 



16 



