1 91 8.] Under Wing-Coverts of Birds. 531 



Practically nothing has hitherto been attempted in the 

 M'ay of particular descriptions of the various modifications 

 of the reversed under coverts in different groups of birds. 

 There is more on this subject in Sundevall's treatise (1) than 

 in any succeeding memoir that I have seen. There it is 

 stated as a universal fact, what will be seen in the sequel to 

 be only generally true, that "on the cubitus the feathers of 

 the first of these two series are firmly attached and just like 

 the remiges, with the inner (posterior) margin free^ covering 

 the outer (anterior) margin of the next feather ; but in the 

 second series they are movable, and lie with the margins in 

 tiie opposite direction to the former, so that the outer edge 

 of each feather is free and covers the inner edge of the next 

 oue.^' Sundevall says he never found an exception to the 

 above rule, and Pycraft (3) says the same thing ; and they 

 both add that the invariable overlap is a certain means of 

 telling, when part or all of one series is absent, to which 

 of the two series the remaining feathers belong. The 

 numerous examinations of wings to be described in the 

 following pages clearly show the above view to be wrong ; 

 the overlap is not invariable in either series of the reversed 

 coverts, and is no certain criterion for deciding to which of 

 the two rows any particular feathers belong. 



In the space of two or three pages Sundevall 'describes 

 the characteristic modifications of the reversed under coverts 

 in quite a variety of birds. Of Columba he says that the 

 major under coverts on the hand are ''first interrupted, 

 then again continued," while the median coverts upon the 

 hand "seem to form a single row with" the major coverts. 

 The facts as seen by him agree exactly with those found by 

 me in other Doves, as described hereafter but are differently 

 interpreted — the major coverts should really be regarded as 

 one uninterrupted series on cubitus and manus, which 

 changes its overlap on part of the manus; and the feathers 

 which Sundevall describes as median coverts, " which seem 

 to form a single row with" the major, are really major 

 coverts overlapped the contrary way. Sundevall's slight 

 observations on the reversed under coverts of other birds 



2q2 



