422 Mr. R. S. Wray on the Structure 



the proximal half o£ every barb, since no one barbule bears 

 at any time the whole of any strain placed directly upon it, 

 but the strain is of necessity distributed. 



The distal barbules present a totally different modification 

 (fig. 4). Each one is a thin lamina bearing on its distal 

 two-thirds barbicels, hooked in the centre of the barbule, 

 but not at the end. The proximal third is a flat lamina 

 about as broad as the proximal barbules^ then it presents two 

 or three tooth-like processes (fig. 4, 1) and becomes quite 

 narrow, beariug five hooked barbicels, and then four or five 

 unhooked ones, the end of the lamina being continued as a 

 barbicel-like process. These all occur on its lower edge. 

 On its ujjper edge there occur as many small tooth-like pro- 

 cesses (fig. 4, 4) as there are unhooked barbicels, the two 

 forming four or five pairs of processes. 



If a section of a barb be examined showing the attachment 

 of the barbules, the proximal barbules are seen to be attached 

 somewhat lower down than the distal ; and the proximal 

 barbules slope gently downwards from their attachment, 

 while the distal slope gently upward (c/. fig. 1). The distal 

 barbules overlap the proximal, the overlapping being con- 

 fined to the pai't beyond the teeth (1. fig. 4), the ends of 

 the unhooked barbicels lying over the upper edge of the 

 barb often. The booklets (hamuli) catch the flange of the 

 proximal barbules, and, being regularly arranged, each 

 booklet has a grip on a separate barbule; each distal bar- 

 bule, therefore, is connected with five proximal ones, and as 

 each of these is intimately connected with its fellows by the 

 overlapping previously described, the whole forms a most 

 complete adaptation of structure to function {cf. figs. 1 

 and 4). 



The methods adopted in procuring data for the making of 

 the model were as follows : — 



(1) A feather was soaked in turpentine, and bits of the vane 



were cut out and mounted in Canada balsam, show- 

 ing the upper and lower surfaces. 



(2) Separate barbs were taken and mounted, the barbules on 



some being teased out with needles, on others the 



