1886.] 



CUBITAL COVERTS OK BIRDS. 



187 



species of Passerine birds, and will probably, on further examination, 

 be fonnd to characterize the whole of the birds that are correctly 

 referred to that Order. 



In the Corvidae an approach towards a somewhat different mode of 

 arrangement is made (fig. 3) : another minor modification is seen in 

 the Alaudidse (fig. 3a). The Swallows (fig. 5a, p. 188) all appear to 

 follow the normal passerine type. That of the Swifts and the Cotingas 

 appears to me to be essentially different. There is some doubt also 

 in regard to the Bower-birds and the Birds of Paradise in this 

 respect. 



Following Dr. Sclater's arrangement, the Swifts and the Humming- 



Fig. 2. 



Fig. 3 



Fig. 3 a. 



I I, \ \ \ 

 Turclus mcrula. 



Corvtff: 



Alainla arvmsis. 



birds fall next to be described. Living Humming-birds can very 

 rarely be examined closely ; I have therefore been compelled to rely 

 entirely upon the examination of museum specimens. After exa- 

 mining the whole of the Gould Collection, and checking the results by 

 comparing them with those made on a large series of other specimens, 

 I am convinced that one general type of wing-pattern characterizes 

 the whole of these birds ; it is of a very simple character, and is 

 represented in figure 4, p. 1 88. By this it will be seen that the proxi- 

 mal lapping row of median coverts found throughout all the Passeres 

 is absent entirely in this. The Humming-birds might, indeed, be 

 described as possessing no median coverts at all, the place of these 

 being taken up by feathers having the same mode of imbrication as 

 the Lesser Coverts. All the feathers of each series overlap outwards 

 and backwards from the vertebral axis towards the distal end of the 

 wing in these birds. 



Observations on the order of overlap in the wing of freshly-killed 

 specimens of Cypselvs apits, afterwards extended by an examination 

 of the whole series of Swifts in the National Collection, showed that 

 in these, as in the Humming-birds, no one series of feathers overlaps 

 backwards. In fact the wing-pattern in the genera Ci/pselus, 

 Acanthylis, Chcp.tura, and Collocalia seems to me to differ in no 

 essential respect from that foimd throughout the Trochilidse. So 

 far as the disposition of the wing-coverts is concerned, the Swifts and 

 Humming-birds agree amongst themselves, and differ from all of the 

 Passeriform birds, with the possible exception of the Birds of Paradise. 

 Fig. 5, p. 188, taken from a freshly-killed specimen of Cypselus apus, 

 will serve to make this poiiit clear. A wing of Hirimdo nistica is 

 figured alongside for comparison (fig. 5 a). 



L3* 



