J. E. DUERDEN 145 



ordinary purposes it may be held that the northern and southern birds 

 have the same average of plumes. By selective breeding numbers of 

 chicks bearing more than 40 plumes are now being reared. 



Whether degeneration is in progress among the remiges of the ostrich 

 at the present time can only be indirectly established, for no standard 

 exists with which the number in the row can be compared. In the case 

 of the coverts any loss is at once apparent, since the normal alternations 

 with the remiges are wanting. The wide variations from 44 to 33 might 

 possibly be regarded as the normal fluctuation of the mixed race of 

 ostriches, and it will be shown that each plume has distinct factorial 

 representation in the germ plasm. As in fluctuating series generally few 

 individuals are represented at the two extremes, but more are collected 

 about the middle. To establish degeneration statistically it would be 

 necessary to show that the average for the whole race is undergoing 

 retrogression, but this is manifestly impossible. 



Comparison of the wings of high and low numbered ostriches reveals 

 that losses are eflfected at the elbow end of the row. In high numbered 

 birds the remiges extend round and beyond the elbow projection, and 

 pass for a short distance along the upper arm, an unusual relationship 

 in birds generally ; but in low numbered they scarcely reach to the end 

 of the elbow, and a wide interval separates them and the first rows 

 of upper-coverts from the rows of the upper-arm, rendering counting 

 much easier (cf Figs. 3 and 4). Occasionally one or two diminutive or 

 vestigial feathers occur at the termination of the row, similar to those 

 found at the ends of the covert rows, and these may be regarded as satis- 

 factory testimony that reduction is actually in progress, particularly 

 when taken in conjunction with the facts of degeneration generally. The 

 process must however be very slow and fairly uniform for the race as a 

 whole, free from anything fortuitous, seeing that both the northern and 

 southern birds bear the same average. It may be held that the ancestral 

 ostrich would have a fixed number of wing-quills at the beginning, and 

 that any later departures fi."om this would be in the direction of reduc- 

 tion, not of increase ; hence we may regard all losses below 44 as so 

 many losses of individual plumes and as retrogressive mutations. The 

 later embryos show the same variations as the adults, a proof that the 

 germinal factors concerned with the lost plumes have disappeared from 

 the germ plasm, following the same ordinal sequence from the proximal 

 end outwards. 



None of the ostriches in Africa are yet germinally pure so far as 

 concerns the number of wing-quills. The crosses show them to be 



