582 On Reversed Under Wing-Coverts of Birds. [Ibis, 



disappearance of the cubital major under coverts, but so 

 irregularly that no series of families or genera can be formed, 

 since great variation is found within genera and even species, 

 some individuals having the rudimentary major under coverts 

 quite large, while others of the same species have them very 

 small or even absent. Yet in the Swallows these rudi- 

 mentary coverts were always present, and also in some small 

 groups of which but few representatives were seen ; and in 

 the small African " Warblers " related to Cisticola, and 

 in the Sunbirds, they were nearly always wanting. 



How far phylogenetic relationship may be inferred from 

 such series of groups of Inrds as those given above is a diffi- 

 cult question. Of course it is quite impossible to suppose 

 such relationship between some of the groups placed together 

 in the above series, as between Doves and Coraciiform birds. 

 But this fact does not destroy the force of the evidence of 

 the reversed under coverts in favour of relationship in other 

 cases. That the characters derived from these coverts can- 

 not be used in all cases with logical precision is merely what 

 must be admitted of all characters used in classification 

 whatever. Furthermore, where so much uniformity of type 

 is found in the reversed under coverts in undoubted large 

 groups of birds, a departure from this uniformity is a real 

 ground for doubt about the inclusion of some groups in the 

 larger grou[)S in which they have been sometimes placed. 

 The Owls and the Nightjars, for instance, liave the reversed 

 under coverts little modified from the primitive or "normal" 

 type, and not showing the slightest tendency towards the 

 very peculiar and characteristic type seen in the Picarian 

 birds (Kingfisher-Woodpecker group). So also the Parrots, 

 with the two rows of these coverts well developed, show no 

 tendency towards the type found in Plantain-eaters and 

 Cuckoos, in which the median coverts have disappeared, 

 though the Fowls, with their much reduced median coverts, 

 do show such a tendency. It may be added that the reversed 

 under coverts of the Swifts, in the few examples seen, do 

 not appear derivable from the type found in the Kingfisher- 

 Woodpecker group — or from any other special type observed. 



