118 J. Christopher Haney et al. 



k* 



Fig. 1. Specimen of nominate Pterodroma mollis, showing underparts, collected 

 from Antipodes Island, 13 February 1969, by J. Warham (Department of 

 Zoology, University of Canterbury, Christchurch 1, New Zealand). 



Madeiras (James and Robertson 1985). Measurements of wing and 

 body length in Fea's petrel overlap extensively with the black-capped 

 petrel (Cramp and Simmons 1977, Harrison 1983). When we (J.C.H. 

 and C.A.F.) compared specimens of P. feae and P. hasitata, we con- 

 cluded that size differences between these two species would be difficult 

 to detect under field conditions. The Freira petrel (a summer breeder) 

 is significantly smaller than either the Fea's petrel or the soft-plumaged 

 petrel (Cramp and Simmons 1977, Fisher 1989). Bourne (1983a) stated 

 the P. madeira has less mottling on the flanks than P. feae. However, 

 other accounts differ (see Cramp and Simmons 1977, Harrison 1983), 

 and such relative field marks may not be useful at sea in the absence 

 of direct comparison and experience. 



Status and distribution — We believe that the bird we observed 

 must have been P. feae for the following reasons. First, the lack of 

 a complete breast band eliminated southern soft-plumaged petrels. None 

 of >290 specimens of the northern forms or numerous living examples 

 examined by P. and F. Zino (personal communication) have a marked 

 breast band, although a small minority of individuals with breast bands 

 have been reported at sea off Madeira (taxonomic identity unknown; 

 B. Zonfrillo personal communication). Generally, there is at least 1 

 cm of white between the two sides of the partial breast band in the 

 northern species (F. Zino personal communication). Second, the large 



