MR. E. DEGEX OX ECDTSIS. 397' 



part of the spinal tract, and it affords me great satisfaction to have independently 

 arrived at the same result as that of Mr. Dwight, whose paper I did not see until towards 

 the close of my work. 



Of these, one proceeds from the lumbar, one from the thoracic, and one from the 

 interscapulary region (this latter beginning to moult later: see Chart, Table VI.). lu 

 addition to these main centres, there is to be noted another for the portion to which 

 I refer as " the Lower Neck Region " in the Chart. I have also pointed out the lateral 

 rows as such, from which moult proceeds later, as a secondary feature. 



I am also able to state that I had a similar notion about the agency of the blood- 

 supply possibly influencing the early moulting conditions of these centres, as referred 

 to by Dwight (/. c. p. 84), to which the specimen No. 12 testifies, and from which 

 the present description of this tract is made, by having used a carmine injection for the 

 purpose. In this I failed to procure the desired evidence, on account of some 

 unsuspected internal injuries by shot. 



From these main centres the moult spreads both forward and backward simul- 

 taneously, but slightly more in proportion in the latter direction. It shows signs of 

 energetic progress nearly always in advance of any remiges, as shown in my earliest 

 acquired specimens. It is also protracted during and often beyond the latest phases of 

 development of flight-feathers, as specimens in which the moult of the latter is 

 completed show. The middle row is in full vigour of new growth, and approaches 

 completion much sooner than the time of appearance of primaries III. -IV., as shown 

 on the Chart; the crown of the head, and notably the auricular regions (lores, &c.), 

 being the latest of the lateral rows to change, except the sides of the nape. 



The Humeral or Sca'pulary Tracts (Pterylte humerales). 



Pterylographically, this tract, which is a double one, has been defined by Nitzsch 

 (37, p. 25) as being entirely divided by space (the lateral Apteria spinale) from 

 the spinal tract previously considered. Although being more directly connected with 

 the " inferior tract " by means of a band of closely-set rows of feathers, it is linked, 

 nevertheless, in this particular species to the " spinal tract " by a contiguous set 

 of softly-textured feathers consisting of about three rows. These are placed across the 

 clavicular region in front of the head of the humerus, thus bisecting the long spinal 

 apteria, a similar arrangement as that ascribed to some birds by Nitzsch (p. 25, /. <?.), 

 e. g. Steganopodes, Buceros, and Colius, which, as he remarks, " may nevertheless be 

 recognized as distinct tracts by the stronger formation of their feathers." 



The humeral tract may be subdivided into two distinct portions, each with a well- 

 marked centre from which the moult spreads independently, a humeral and a clavicular, 

 or more marginal portion, the latter merging into the alar marginal tract. Both these 

 consist of several rows after their bifurcation. Of these, those medially situated aro 



3k2 



