EVOLUTION OF OPERCULUM IN SERPULA. 53 



already known concerning the homologies of the branchiae and the 

 operculum in Serpulids and their near neighbours, the Sabellids. 



A. Firstly, from an examination of the nerve supply to these 

 organs, Quatrefages (2) found " that in the Sabellinae as in the 

 Serpulinse, the branchial nerves arise directly from the cerebral 

 ganglia as antennary nerves " (i.e. in the position which such nerves 

 have in such forms as Nereis) " and among the Serpulids a^slemier 

 branch springs, on each side, from the principal trunk and leads to 

 the base of the two opercula, as well to the rudimentary one, as to the 

 one completely developed. The branchiae and the opercula therefore 

 correspond to antennae." Pruvot (3) further shows that the antennas 

 modified into gills in Serpula, are the second pair, those large 

 organs that in Nereis, Polynoe, &c., are known as palpi. 



B. Embryology shows that very early there appear (Sabella) 

 two ciliated wing-like processes on the dorsal aspect, which shortly 

 after divide each into two slender lobes which constitute the first 

 four branchial rays, their number augmenting rapidly by budding 

 on the ventral side (4). 



C. Next examining the near relatives of Serpula, we find in 

 Sabella no opercula, but in place, there are two slender, simple dorsal 

 processes occupying the same relative position and having the same 

 innervation or nerve-supply as the opercula of Serpulids whence we 

 may justly infer that they had similar origin. 



In some Sabellids (Othonia) these simple antennae are wanting, 

 as is the case too with a number of the Serpulids. Thus Apomatus 

 and Filigrana possess only branchiae and to make amends for want 

 of a true operculum as in the more typical species, in these the 

 summit of one or more of branchiae which are of the usual bipinnate 

 style is enlarged in a bulbous manner into a pseud-operculum. 



D. Finally Fritz Mliller (5) has described a Serpulid which 

 when first observed had but three pairs of pinnate gill-filaments, but 

 which in a few days time, he found had developed a clavate oper- 

 culum at the extremity of one of the filaments. In the course of 

 three days more a new pair of branchial filaments had sprouted forth 

 and the opercular peduncle had lost its lateral filaments. 



So much for collateral evidence. Now let us see if we can 

 gather the threads together, and make a web of fairly good strength. 



From the frequent assumption of opercular shape by the or- 

 dinarily non-functioning and simple antenna in Serpula, and from 

 its inconstancy to one particular side, we may justly infer that this 

 variation is atavistic, and therefore that at one time, Serpula pos- 

 sessed two functional opercula. Then, from the corresponding 

 relative position of the organs, we have already inferred that the 



