2 
The nature of this relatively straight point-to-point survey, covering 
vide latitudinal and longitudinal ranges, makes the data conducive to inter¬ 
pretation as a chainlike series of events. In the same way that a length of 
string can be cut up more ways across its axis than parallel to it, much of 
the data in this cruise is best presented in serial relationships showing 
variations along a line. A series of Tables has been constructed to demon¬ 
strate how this approach can express changes and other biological patterns 
of the birds under study. 
In general, the tables presented in this appendix differ somewhat from 
most tables in surveys which cover a more restricted area or track. A dis¬ 
cussion of each table should clarify the meaning and significance. 
TABLE 1. Summary of Daily Observations 
In order to simplify other tables, each of the 13 days on which regu¬ 
lar observations were made, was assigned a number (l-13). Note that no 
regular observations were held on 6 June; this was due to illness of the 
observer. The mean linear density is high due to the influence of day #7. 
Only two daily linear densities exceed the mean. The'median daily linear 
density of 5.4 birds/linear mile is perhaps a better measure of the central 
tendency. The "approximate distance to land" column is relevant to the 
"major species" column: above 75 miles major species are Storm Petrels, 
Albatross, Phalaropes, and Gadfly Petrels, with Terns, Boobj.es and Shear¬ 
waters found closer to land. 
TABLE 2. Relative Abundance of Birds by Species 
All species were not seen each day, therefore the overall linear den¬ 
sity for any species is lower than the actual density for the area in 
which the species was present. The average linear density for days of oc¬ 
currence is obtained by dividing the total birds seen by the total miles 
for those days that the species was recorded. The quotient obtained by 
dividing the occurrence density by the overall density would give a measure 
of clumping or concentration tendency. Thus, a restricted form such as 
Cooks Petrel measures about 5 units on an artificial scale whereas 
"White-rumped" Storm Petrel measures approximately 1 unit. 
TABLE 3 presents a similar picture on a higher taxonomic level. 
TABLE 4. Species Contributing or More to Total Birds 
This table gives a quick overall view of relative concentrations. 
Comparison with the occurrence rank again shows a relationship with clump¬ 
ing tendency. 
TABLE 5. Regularity of Occurrence 
This table is an expression of occurrence and measures both breadth 
of distribution or extent of range and regularity along the track. 
