were recorded and 21 birds of 6 species were collected. Throughout the grid cruise 
seas were too rough to use the small skiff for collecting purposes. However, 
collecting was deliberately kept to a minimum throughout most of the survey because 
of more fruitful results obtained in the form of streamer sightings. Invariably 
flocks could only be approached once before scattering. During this brief approach 
the observer must decide whether to attempt collecting or spend the time quickly 
glassing as many birds as possible in search of streamers. The second course 
was usually followed and a total of 33 blue-streamered Sooty Terns were recorded 
during the six day survey (Table 5). Most of these sightings were made on 17 
August W-SW of Howland. This area is not considered to be frequented by the 
apparent majority of Howland birds, but rather all conditions(following seas, wind 
and light) were favorable for approach to and surveylance of flocks throughout 
the morning and early afternoon of 17 August. 
Nocturnal observations were maintained from sunset to sunrise or approach to 
islands. A total of 5UU miles and 59 hours were completed during the grid period. 
Only 232 birds of 10 species were recorded, the majority of these were Sooty Tor ns 
upon close approach to Howland Island. The pattern appeared similar to that 
encountered during July in that the vast majority of Sooty Terns returned to the 
island at dusk from all points of the grid and left during early dawn. No birds 
were collected during nocturnal operations. 
Density (birds/linear mile) dropped to 7.68 from a high of 10.39 recorded 
during July. This is considered a direct result of comparably less survey time 
spent in close proximity to Howland Island during the present cruise. Flocks 
encountered throughout the survey were entirely of Sooty Terns, frigates, and 
boobies. Sooty Terns represented 83*6$ of all birds in flocks (Table 6); not a 
