common or uncommon, January-April. P. e^. 

 cervicalis was probably an uncommon or rare 

 visitor all year. No attempt was made to dis- 

 tinguish between these similar forms until 

 October 1964. 



P, externa total 



Months 



Year 



1964 



1965 



Months 



Year 



1964 



1965 



January 



February 



March 



April 



May 



June 



July 



August 



September 



October 



November 



December 



X 

 X 

 X 

 X 

 X 



X 



4 



4 

 13 



1 



2 







14 



12 



Both subspecies of Pterodroma externa were 

 found in the study area. Even when the pres- 

 ence of both subspecies in the central Pacific 

 was confirmed by the collection of specimens. 

 It was possible to distinguish between them 

 only under conditions favorable for observation. 

 Frequently no assignment to subspecies was 

 possible. Only 2 percent of all P^. externa ob- 

 served were assigned to cervicalis subspecies 

 (largest proportion 16 percent in January 1965). 

 The seven specimens collected in the study 

 area were all P. e. externa . 



Monthly fluctuation of numbers : Monthly totals 

 from March 1964 through September 1964 in- 

 cluded records of P. e. cervicalis . It is prob- 

 able that a small percentage of P. externa , 



identified at a distance as P. e. externa (which 

 come from Mas Afuera In the Juan Fernandez 

 group, 500 miles west of the Chilean coast), 

 were in reality P. e. cervicalis from the 

 Kermadec Islands, 600 miles northeast of New 

 Zealand, and, to a lesser extent, vice versa. 



In February the species was represented only 

 by a straggler or two. Numbers Increased 

 slightly in March and April, and rapidly in May. 

 A nonbreeding season plateau was reached in 

 July and maintained until October, when num- 

 bers reached their peak. Thereafter numbers 

 declined rapidly each succeeding month as the 

 birds migrated out of the study area back to the 

 breeding grounds. The low point was reached 

 at the same time that eggs were beginning to 

 hatch on the breeding grounds. 



Distribution : The species' distribution in the 

 study area reflected Its breeding phenology. In 

 the months in which the species breeds in the 

 Southern Hemisphere. P. externa was found 

 mainly in the southeastern corner of the area 

 (figs. 6a, 6b). During these months (February- 

 April) not only was the range of the species in 

 the area the most limited but numbers were 

 also lowest. From May onward the limit of the 

 range was pushed northwestward until Septem- 

 ber and October, when the species was found 

 throughout the area. Its range contracted pro- 

 gressively in November and December. The 

 distribution in January did not follow the same 

 pattern: although numbers were greatly reduced 

 from the previous month, birds were found 

 throughout most of the area. Since the breeding 

 season was already underway, these birds were 

 probably nonbreeders spending at least a por- 

 tion of the breeding season In their "wintering" 

 area. January was the only month that lacked a 

 defined center of density. In all other months 

 greatest densities were in the southeastern 

 portion of the area. 



P. e. cervicalis numbers and distribution : 

 Although data on this subspecies were Incom- 

 plete, their numbers tended to parallel those of 

 the species in general. February, March, and 

 April were the low months for both subspecies. 

 Owing to the small sample size, as well as the 

 likelihood of confusion In distinguishing between 

 the two forms of P. externa in the field, dis- 

 tributional differences cannot be determined. 

 In general, P. e. cervicalis seemed to have a 

 more northerly distribution only in December 

 and January. In all other months the distribu- 

 tion of the two forms was apparently coincident. 



