J. E. DUERDBN 135 



are domesticated birds the change to this state from the wild one is so 

 recent that it is impossible to conceive that retrogression of such a deep- 

 seated nature as here described could have been instituted in so short 

 a time, at most not more than fifty years ; moreover, the imported 

 northern birds were caught as chicks from the nests of wild birds, and 

 display similar stages to those of the domesticated Cape bird. 



Under-Coverts. 



First Row. In nearly all ostriches the under surface of the wing is 

 naked except? for the presence of a single row of under-coverts, and even 

 this is rarely complete. The coverts alternate with the row of wing- 

 quills, and also with the last plume of the first, second and third row of 

 upper-coverts which, along with the former, constitute the post -axial, 

 marginal row of plumes (Fig. 1). Only in two birds out of hundreds 

 examined however has the full alternation of under-coverts been ob- 

 served, and often as many as nine or ten are missing. Actual counting 

 of twenty-five birds gave an average of 30"2 coverts, which is six less than 

 the average of the remiges, and one, two, or three more should be present 

 for alternation with the last member of the one or more rows of upper- 

 coverts completing the marginal row. The absent members form a 

 regular ordinal succession, beginning at the elbow end of the row, and all 

 numerical stages from one to ten can be procured, though the greater 

 number of birds have only six or seven missing. The remiges constitute 

 a convenient standard from which the degi-ee of loss of the coverts can 

 be determined. They have a fluctuating variation of their own with 

 which that of the coverts is correlated, but the latter have also an inde- 

 pendent variation. 



While usually the losses represent entire plumes, not infrequently one 

 or two vestigial feathers are seen beyond the last ordinary one. These 

 are sometimes merely diminutive entire feathers, but often they are 

 small and imperfectly formed, the shaft especially being wanting. The 

 stages of degeneration which they represent are similar to those more 

 frequently found constituting the degenerative down on the wings and 

 tail to be described later. Here all stages in the reduction in size of 

 the complete feather are to be seen; then others in which the shaft 

 breaks up into two or three imperfect parts and ultimately disappears, 

 and only a tuft of barbs bearing barbules results; afterwards the barbules 

 disappear, and, lastly, one or two hair-like barbs only remain (Fig. 2). 

 These degenerate feathers continue to appear in the individual with the 



9—2 



