120 Clark, Tail-feathers and Coverts. LApril 



Pelargiformes. In Ibis, we find 6-6, and the same is true of 

 Ardea, Herodias, Nycticorax and Tantalus. Cancroma shows 6-7, 

 the extra covert being above rectrix 6. Nitzsch says that he found 

 only ten rectrices in Cancroma. I had but one specimen and it is 

 unidentified but there are clearly twelve rectrices, the number 

 accredited to Cancroma in the ' British Museum Catalogue.' The 

 coverts alternate with the rectrices as a rule. In Botaurus the 

 formula is 5-5 or occasionally 5-6, suggesting that the reduction 

 in the tail of this genus is quite recent, the extra covert being a 

 relict from the former condition of 6-6. The coverts in the bittern 

 are above the rectrices. 



Phoenicopteriformes. I was fortunate in being able to 

 examine four flamingos. In three, the arrangement was 7-8, but 

 in the fourth it was 6-7. The last specimen had a normal tail 

 but with only twelve rectrices. The first covert is the longest and 

 lies outside the first rectrix. 



Anseriformes. The single swan examined showed 10-12, the 

 extra coverts outside the last rectrices and by no means clearly 

 distinct from other contour feathers. In Branta, I found 8-9, 

 and also in Anas. In Spatula, Cosmonetta and Nyroca, the formula 

 is 7-8 or 9. In Clangula, we find 8-9 or 10 and in Dafila, 8-10. 

 In Erismatura, 9-10 or 11 occur. In all Anseriformes, we find then 

 more coverts than rectrices. In general each covert lies over its 

 rectrix and the additional coverts are at the outer end of the series, 

 where they are often hard to distinguish from the ordinary contour 

 feathers. Occasionally covert 1 lies beside rather than above 

 rectrix 1. In Clangula, the series of major coverts is nearer the 

 end of the pygidium at the middle than it is on either side; in other 

 words the covert-series is not parallel to, but divergent from, the 

 rectrix-series. 



Pelicaniformes. In all steganopodous birds, except Plotus, 

 the number of coverts corresponds to the number of rectrices; the 

 same is true of their position as a rule. But in relative size there is 

 more diversity. In Phalacrocorax, we find 6-6, with the extraor- 

 dinary condition of the coverts described above (p. 115). In Sula 

 we find 6-6 or 7-7, coverts and rectrices corresponding in position 

 and size. In Frcgata, there are 6-6, with coverts 2 and 3 the 

 longest but 1 not much smaller. In Phocthon, I find 7-7 or 8-8, 



