J. E. DUERDEN 141 



of a continuous series of variations from one extreme stage to another 

 does not necessarily represent a successional germinal series, nor indi- 

 cate an evolutionary trend in any determinate direction. In his studies 

 on Drosophila Prof Morgan, for example, shows that a continuous series 

 of intermediate eye-colours between the two extremes, deep brown sepia 

 and pure white, could be picked out ; also a graded series with perfectly 

 formed wings at one extreme and no wings at the other. But the experi- 

 mental evidence reveals that the various stages are produced as separate 

 mutations, independently of one another, not as a sequential series, and 

 that large steps occur as well as small ones. Detached changes of this 

 kind are not steps in evolution unless they confer selection value upon 

 their possessors, but are examples of the small, fortuitous germinal varia- 

 tions probably always to be found in a mixed assemblage of any species. 

 In the ostrich, by contrast, everything indicates that the variations are 

 successional in their origin, that all have a retrogressive trend, and are 

 a part of the evolutionary process common to the race; the different 

 individuals represent the different evolutionary stages of the race. From 

 the evidence here presented no one can doubt, for example, that we have 

 a definite sequential trend in the loss of the toes from the foot ; also 

 in the loss of the coverts from the wing, the one or two diminutive or 

 vestigial feathers at the end of a retrogressive row being taken to repre- 

 sent the magnitude of the losses in progress for the time being. It is 

 claimed that the ostrich is in a retrogressive evolutionary phase at the 

 present time in various directions, and that the different stages repre- 

 sented by the different members of the race afford us the means by 

 which the methods followed can be determined. 



Upper-Coverts. 



The upper-coverts are black in the cock and grey in the hen and, like 

 the under-coverts, regularly alternate with the wing-quills (Figs. 3 and 

 4, and PL V, fig. 1). Usually only the first two rows are clipped as part 

 of the feather crop. The complete number of rows however varies in 

 different individuals. In some only two or three are clearly defined 

 (Fig. 4), but five or six in others, or even more if the counting be con- 

 tinued to the short rows near the pre-axial border of the wing (Fig. 3). 

 As regards the number of plumes to each row, the first one is found to 

 vary in direct correlation with the wing-quills and rarely exhibits any 

 independent reduction. The first member alternates with the first and 

 second wing-quills ; the last one extends beyond the last wing-quill and 



