Respiration in Invertebrate Animals. 305 



vertically over the whole group of gills. From its position and 

 structure, it can affect only the outermost surfaces of these organs. 

 If this instrument were a simple, flat, smooth-edged process, it 

 is easy to conceive, that during its motion over the branchiae 

 from above downwards, the component tubuli of the latter could 

 not by mechanical possibility be separated so as to favour the 

 rush between them of the aerating element. This difficult pur- 

 pose is accomplished by an inimitable contrivance. The flabellum 

 is covered in rich profusion with minute, flexible, mop -like 

 threads (fig. 8, E) . At the distal extremity of each seta or bristle, 

 a group of minute flexible processes are added. A structure of 

 so great singularity cannot be misinterpreted. The purpose 

 which they are designed to fulfil cannot be mistaken. They 

 constitute artfully-adapted provisions for cleansing, mopping, 

 separating, agitating the constituent filaments of the branchiae. 

 No other description of instruments would answer the same ends. 

 They are not merely substitutes for cilia. Cilia would simply 

 effect the rapid renewal of the aerating element. They answer 

 the manifold uses enumerated. Another variety of seta is inter- 

 mixed with the former on the same flabellum. It consists of a 

 sword-shaped process (fig. 8,/), less flexible than the former, from 

 the edges of which secondary, acute, minutely-delicate points 

 arise at an obtuse angle. They are designed only to act in one 

 direction. They are situated chiefly on the margins of the fla- 

 bellum (D). They exist on those placed vertically between the 

 gills, as well as on that long whip-like flabellum which acts hori- 

 zontally in the branchial cavity. They are less fitted to wipe the 

 surface of the tubules than to catch at their edges when swiftly 

 drawn over or between them. Thus they separate and mo- 

 mentarily hold apart the slender filaments of the branchiae. 

 Nearer to the root of each of these setae, and only on one side, 

 a second system of angular teeth occurs, which are turned back- 

 wards towards the root (PL XVII. fig. 9, C, E, D). They are thus 

 capable of acting in a direction the reverse of the serrations placed 

 on the extremity of the same seta. The mop-like variety of setae 

 does not exist on the flabella of those Crustacea the branchiae of 

 which are leafy or laminar. 



Under the latter circumstances they would prove ineffective 

 as cleansing utensils. They could perform no mechanical work 

 on passing between smooth parallel laminae. Between cylin- 

 drical filaments they act far otherwise. Nothing can surpass 

 the efficiency of the contrivance. There exists such an evidently 

 suitable relation between the structure and form not only of the 

 flabella, but of their minutest hairs, and the shape and figure of 

 the branchiae in the same subject, that the incomparable inge- 

 nuity with which these little implements are adapted to the exact 



Ann. ^ Mag, N. Hist. Ser. 2. Fo/. xiii. 20 



