44 
of the relationships existing between bird distribution, vertical tem¬ 
perature, structure of the water, and latitude. Eor the most part other 
possible parameters such as weather, time, proximity to land, and longi¬ 
tude are not dealt with or are mentioned only in passing. Temporal varia¬ 
tions in bird distribution are very important in this region, but are not 
analyzed in detail here because the four-month period involved is, on the 
whole, a fairly static part of the years. 
Water components vs. latitude 
«r 
Since the oceanographic features of this region of the Pacific con¬ 
sist primarily of east-west belts, the horizontal rate of change is very 
great with respect to latitude and very small with respect to longitude. 
I have little information on longitudinal variations of the features to 
be discussed but my feeling is: for the oceanic area between 0° - 20° N.; 
l40°W- 180° variations in the east-west direction at any given time are 
negligible compared to the seasonal change. Consequently, the effect of 
longitude is not considered further in this treatment. 
♦ 
There are four major water masses crossed when traversing the some¬ 
what over 20° of the latitude between the Equator and Oahu, l) South 
equatorial current — up to ca. 3° N. 2) Equatorial countercurrent — 
ca. 3° - 12° N. 3) North equatorial current -- eg,. 12° -l6° N. 4) North 
Pacific centrals water north of 1 6 ° N* 
Data Analysis 
Bathythermograph (BT) data were taken during all or part of each 
cruise through the region. While the data have been sent to other organi- 
cations for final analysis, some preliminary analyses have been done dur¬ 
ing and immediately following the field work by Dr. Robert Pyle and myself. 
In many respects the preliminary analyses prove to oe more valuable xor at 
least 2 reasons, l) The data can be extracted immediately and used to 
supplement current bird observations. 2) Pertinent measurements of tne 
data can be selected in order to best show relationship to the bird ooserva- 
tions. The BT data and bird distributions are both presumed to show some 
degree of association with the availability of suriace iood. We are inter¬ 
ested in relating the BT information to the bird distribution because the 
3T measurement is more readily obtained than any other measure ox the .otu 
availability. Viewed in this light both the bird observations and the ^BT 
soundings become a set of parameters measuring a single phenomenon, and 
can be treated rather artificially. The goal in concurrently presenting 
the two sets of data is to demonstrate where objective similarities vP e nks, 
lows, etc., on a graphical display) exist, and to interpret the correla¬ 
tions subjectively in light of their ecological significance. 
\ 
A series of such manipulations have been made and presented graph¬ 
ically and some appear useful. They are presented after a general section 
discussion. 
Bird Sections 
i 
A cursory inspection of the data consistently shows that two peaks 02 
/ 
