are suspected of considerable annual migrations, which it was thought, when the scouting pat- 

 terns were planned, might be detectable by a "Doppler effect", this assumption may appear 

 unjustified. However, our data, which will be presented in the section on migration of skipjack 

 schools, indicate that no such effect was detectable and that the movement of schools while being 

 watched approximated a random motion. 



Another necessary assumption which appears to be well justified is randonn distribution 

 of schools in the scanning area. The totals seen in the starboard and port forward quadrants on 

 runs in the various directions during GILBERT cruises 11 and 13 are shown in table 4. The data 

 suggest a tendency toward more sightings in the starboard quadrant, but the difference between 

 starboard and port quadrants is not statistically significant by a chi-square test. The total num- 

 ber of flocks seen according to the relative bearing from the bow did show a significant tendency 

 to increase at bearings between 30 and 60 during the spring scouting (table 5). However, this 

 appears to be associated with characteristics of the vessel and habits of the fishermen. The 

 GILBERT has a rather high bow and a mast and shrouds forward of the bridge, where the ob- 

 server stands, which tend to obscure vision directly aheeid of the vessel. Also, the fishernnen 

 perhaps pay somewhat less attention to the area directly ahead of the vessel, because they believe 

 that any flocks in that area will be approached by the vessel and be seen eventually. Quite likely 



these two factors resulted in a lesser number seen directly ahead of the vessel and a somewhat 



o o 



greater number in the 30 to 60 sector. 



Table 4. 



-Number of flocks sighted in forward quadrants by course run on GILBERT cruises 

 11 and 13-' 



— A total of 5 flocks (3 seen in after-quadrants, 2 with incomplete data) omitted from this 

 tabulation. 



Table 5.--Nunnber of flocks by relative bearing from bow 



— Starboard and port sides combined. 



Finally, it is necessary to assume that all of the flocks and schools within 1/2 mile of the 

 vessel were seen. Unquestionably this is true of the bird flocks, but the flocks habitually assenn- 

 ble and disperse while accompanying fish schools, and the assumption that all of the fish schools 

 which approached within a half-mile of the vessel were accompanied by bird flocks at that parti- 

 cular time may not be justified. The ease with which a considerable flock of birds is seen at a 

 distance of 2 miles is not compatible with an estin-iate that an average of only 5. 9 percent of the 

 flocks are seen beyond that distance. Unquestionably the birds are missed to a major extent 

 because they are not assembled into a flock and behaving as they do when they are chasing a fish 



10 



