MR. E. DEGEN ON EODYSIS. 371 



though in the case of Nos. 28 and 29 the cubital remiges IV. and V. were still 

 somewhat short of their ultimate lengths. 



B. Remarks on the Period of Groioth and the Development of the Remiges. 



Although the actual time required for the production of each quill itself cannot be 

 followed with precision in the dead animal, neither is such the case with a bird in 

 captivity, owing to altered conditions. But this can, nevertheless, be estimated tolerably 

 well from a number of specimens which show a particular quill to be in moult within 

 a given stage ; and, further, with the aid of a quill just recently developed to full growth, 

 as is the case in the majority of individuals showing more than one quill in successive 

 stages of development at the time, and another one perhaps just emerging, when we 

 get a valuable means for obtaining, by differential inference, an approximate result for 

 the individual. 



From the averages extracted for the period of growth for each quill, the result shows 

 that those of the earlier stages develop faster than those of the later ones, which take 

 somewhat longer to grow. The difference, however, is not considerable, and fluctuates 

 between seven and ten days respectively. The range of time for the moult of the 

 different remiges during a seasonal change can also be computed for the species as a 

 whole, and a chart (Table II.), setting out the general result, having the additional 

 advantage of showing their synchronic movements, has been appended also at the end 

 of these remarks. 



C. The " Bastard Wing." Alula or Ala spuria = Plumse pollices. 



In all robustly and normally developed individuals of this species I have never 

 failed in finding the regular number of penncB, which constitute this series, to be less 

 than four. It is in weakly developed individuals only that their number is occasionally 

 reduced to three. In such cases it is invariably at the expense of the innermost, or 

 the first one if one regards them in the same light as the metacarpo-digital remiges. 



Had there any additional proof been wanting on the remigial nature of these 

 feathers, apart from a previous observation of the most proximal ones being sand- 

 wiched in between an abortive upper and a homologous ventral covert, nothing could 

 have been more fortunate for dispelling the doubt as to their true value as flight- 

 feathers than that shown by their moult. 



When I had received specimens representing the progressive stages of the moult 

 for a consecutive number of weeks without having met as yet with any evidence on 

 this point, I began to entertain serious misgivings of having missed what I then 

 considered to be one of the most desirable points in this investigation. 



As these feathers, it seems but reasonable to assume, are not greatly exposed to 

 wear, it is at all times difficult, as I found, to distinguish between their old and 

 their new growths, unless one secures the bird during their actual renewal. 



VOL. XVI. — PART VIII. No. 4. — May, 1903. 3 g 



