On the Origin of Feathers. 215 



This is the stage which is the fii'st to appear in the ontogeny 

 of Neossoptiles on the typical scale-less skin of the bird's body. 



4. Papilla ,,sinking in", with the formation of a follicular pocket. 

 Epitrichium still complete. The solid horny cone is further 

 down supplanted by the hoUow sheath and the Neossoptile 

 rami. 



5. Neossoptile and Teleoptile or first final feather. Epitrichium 

 severed by the growth of the Teleoptile. 



6. Three branches of a Neossoptile combining as a spool*, which 

 in turn splits again and is continued as two branches of the 

 final feather. 



7. A Hair for comparison. Epitrichium apparently no longer 

 repeated; Sheath severed or pierced by the growth of the hair. 



It can not be emphasised enough that the whole pulp is not 

 the homologue of the core or body of a scale, except in so far as 

 it represents some of the latters blood-vessels. It is nothing but 

 vascular, a much developed artery and vein wdth much lymphatic 

 meshwork, developed as a consequence of the activity of the 

 düster of epidermal cells. The outer, environmen tal effect upon 

 there cells, their reaction upon pressurt;. Insults, ,,need of pro- 

 tection", is the primary cause; the pulp or swelhng of the meso- 

 dermal parts is a result; and the suppression, not conversion, of 

 the respective portion of the original core is a further result. 



If this prohferation takes place periodically, with inter- 

 vening stages of lesser activity there will result feathers much 

 resembling those which are represented in Fig. 18, PL II of Born- 

 stein's paper; and if the pulp grows much in length, the archaic 

 sohd epidermal cone (cf. Fig. 18) will, in one of the next gene- 

 rations be replaced ly a longer, partly hollow, cone. Such a thing, 

 a cyHnder, closed at the top, is the horny transparent sheath, 

 which encases every growing feather, from Neosso- to Teleoptile. 

 It represents the second stage of the feather' s 

 genesis. The wart-like excrescence, the solid cone, and the 

 Neossoptile are stiU covered the by epitrichium. This archaic, out- 

 most product of the epiderm is, for obvious reasons no longer 

 regenerated from the first Teleoptile onwards. 



The next stage is characterised by the formation of a cylinder 

 within the first, by a repetition of the process of proliferation 

 from deeper strata of the Malp-ghian cells which meanwhile have 

 increased their number of layers. But this second cylinder, owing 

 to irregulär apical growth, is frayed out like a brush; this lowest 

 of Neossoptiles however still recapitulates its ancestral con- 

 d'tion by repeating with its basal portion a solid mantle, the 

 incipient or first spool. 



It Stands to reason that the fraying-out process began at 

 the tip of the whole projecting structure and worked downwards, 



7. Heft 



