202 NATURAL SCIENCE. Sept., 



Owen, writing some thirty years later, gives a much less truthful 

 description. He says, " The barbules arising from the upper side of 

 the barb, or that next the extremity of the feather, are curved down- 

 ward or toward the internal surface of the shaft ; those which arise 

 from the underside of the barb are curved in the contrary direction ; 

 so that the two adjoining series of hooked barbules lock into one 

 another in a manner which has been compared to the fastening of a 

 latch of a door into the catch of a door-post." 



Huxley's description is as follows : — " The barbules may be 

 laterally serrated and terminated by httle hooks which interlock with 

 the hooks ^ of the opposed barbules." 



The last and most authoritative account is that of the late Mr. 

 R. S. Wray. Practically, his paper, short though it is, tells us all that 

 is to be told ; but the author presumably failed to appreciate, or missed 

 entirely, one or two of those nice little differences that will so often 

 escape the notice of the most painstaking workers, of whom he was 

 one. Possibly " 'tis a rule in Nature," in order that those who come 

 after us may have something left to do. 



As just stated, Wray's figures and descriptions may be held in 

 all but a few details to be correct. Exception is taken to his interpre- 

 tation of the proximal radii. These he describes as " . . . thin 

 laminae, with a thickened upper edge forming a small ledge or kind 

 of flange" ; and continues, " If a single barb be examined, the proximal 

 barbules appear ... at their proximal end to join on to an edge- 

 piece running parallel to the barb . . . this edge is seen to be formed 

 by the attenuated halves of these barbules overlapping one another, 

 and being turned at a considerable angle upon the other half." On 

 referring to his figures, however, we find these barbules represented 

 as perfectly flat laminae, vertically inserted by their base along the 

 ramus, and with a slight " flange " running along the upper edge, but 

 with no trace of the "attenuated halves overlapping one another." 

 True, this is a diagram, but surely it would have been more correct 

 to figure the specimen, since the essence indeed of a diagram is 

 to give us the points of a structure, divested of all superfluous detail. 

 So far as my experience goes, this edge-piece occurs but rarely, 

 neither do I believe it is formed as Wray imagined, it, since, in the 

 only case in which I ever found it, the radii were normally related, 

 " the edge-piece," though closely approximated to, being yet quite 

 distinct from them. While this was in the Press I detected a 

 precisely similar flap under the distal radii. Again, as touching 

 the notches, these are accounted for by supposing that they are 

 for the purpose of allowing the distal portion of the radius to be 

 turned upon itself without "crumpling." Would it not seem more 

 probable that these notches are — if we are to explain their presence 

 at all — rather to keep the edges of these radii sufficiently far apart to 

 allow not only of the passage of the booklets, but to prevent their 



^ The italics are mine. 



