EVOLUTION OP THE COLORS OF BIRDS. 185 



about with the tail frequently thrown up in the air. 

 This patch of color would accordingly be of great use as 

 a recognition marking but this subject will be more 

 fully discussed later. 



It thus seems that the markings of the throat, under 

 tail coverts, rump and under wing may be explained by 

 selection. These markings are determined in their 

 shape chiefly, if not entirely, by the part affected, and 

 hence present no particular difficulties. With the head 

 markings the problem is, however, not so simple. We 

 have already seen that lateral markings are common 

 upon the head while transverse markings occur only in 

 special parts and there less frequently than the others. 

 The superciliary stripe is much the commonest of marks, 

 and being generally white in color, there seems far more 

 probability that it is due to some internal principle 

 than is the case with any other bird marking. The fact 

 that frequently where no superciliary stripe occurs in 

 the male it is present in the female or young would also 

 seem to indicate that it is constitutional rather than 

 utilitarian in nature. Instances of this sort are the bay- 

 breasted warbler (Dendroica castanea), red-winged black- 

 bird (Agelaius phwniceus), bobolink (Doliclwnyx oryzi- 

 vorus), etc. There are at least a hundred and six North 

 American land birds in which the white superciliary 

 stripe occurs, from the bob white (Colinus virginianus) 

 at one end of the list, to the wood thrush (Turdus mus- 

 telinus) at the other. In the genus Dendroica it is 

 especially prominent, being present in eleven species 

 and generally very sharply denned. It is also present 

 in nine species of the genus Vireo, and among a large 

 number of warblers of other genera than Dendroica, as 

 well as many of the Fringillid&e. It is not always, although 

 generally, 'a narrow sharply denned line. In the western 

 night hawk (Ckordeiles virginianus kenryi), for instance, 



