22 



THE AMERICAN NATURALIST [Vol. LII 



In some of the split feathers, the division is obviously 

 into right and left halves (Figs, 14, 15, 16) ; in other cases 

 the halves make an angle with each other (Fig. 14, Nos. 37 

 and 80), while in still others one larger part may lie above 

 a smaller part (Fig. 14, lowest row). Whether in the 

 last cases the division has been in a horizontal plane, and 

 in the first cases in a vertical one, is not certain, although 

 the shape of the feathers even in the last case, with the 

 imperfect edge and a narrower margin, would seem to 

 make most probable the view that in all cases the division 



has been into morphological right and left halves. The 

 final position of the feather halves may be due to a later 

 twisting in the sheath, or to crowding of the feathers at 

 the base. This interpretation is further substantiated by 

 cases in which the center of all the feathers has a white 

 area (Fig. 14, No. 55, and Fig. 15) ; this is found on the 

 imperfect side of the split feathers even when they lie one 

 above the other. In all there were 24 F 2 tails with split 

 feathers. Five of these had each two split feathers. 



These cases grade into those in which only the distal 

 end of the feather is split, as shown in Fig. 14, middle 

 figure, and Fig. 16. The impression produced by feathers 



