364 THE ANNELIDA POLYCHAETA. 



the prostomium, though sometimes incomplete above, so as to leave the prostom- 

 ium clearly visible. It bears a pair of often long and conspicuous grooved ten- 

 tacles. It may or may not bear a second pair of shorter processes, often also 

 spoken of as tentacles, but really representing modified parapodia in which 

 traces of setae may sometimes be detected. 



The somites of the anterior division of the body are uniform. They are 

 characterized by having their parapodia all strictly imiramous, the neuropodia 

 not being developed. In the remaining part of the body the parapodia are bira- 

 mous. Each neuropodium forms a bifid, or double, ridge, or torus, on which the 

 uncini are arranged in several series. From differences in the character of the 

 notopodia a median region in most forms is clearly distinguishable from a longer 

 posterior one, the notopodia of the median region often being enlarged and 

 highly modified into fin-like structures. The dorsal surface of the anterior region 

 conspicuously flattened. Dorsally there is a markedly characteristic longitudi- 

 nal median groove, which is cihated and which may extend over the entue length 

 of the body, or may be interrupted. 



The thoracic setae in general are simple lanceolate or hastate forms. The 

 fourth somite always bears one or more stouter setae of a special type. The 

 neuropodia bear numerous seriate uncini which are serrate along their free, or 

 distal, edges. 



Autotomy and regeneration are frequent in the family. Thus, in Chaetop- 

 terus pergamentaceus Cuvier, as a result of unusual stimulus, autotomy takes 

 place between the first and second somites of the median region, and may be 

 followed by regeneration of a complete individual not only from the anterior 

 fragment but also from the posterior one (See Gravier, Ann. sci. nat. Zool., 1899, 

 ser. 9, 9, p. 129, 155). Potts, according to a recent paper (Proc. Zool. soc. 

 London, 1914, p. 955), has found autotomy to be a very frequent and apparently 

 normal method of reproduction in Phyllochaetopterus prolifica Potts. This 

 accounts for the occurrence in the same tube of a number of individuals, a phe- 

 nomenon that had been previously noted by Claparede (Annelides Chetop. 

 Golfe Naples, 1868) for the allied Mediterranean species, P. socialis Claparede. 



The chaetopterids seem never to leave their parchment-like tubes, being 

 among the most sedentary of the polychaetes. The tubes may be straight 

 but connnonly are more or less strongly bent into a V- or U -shape, or sometimes 

 nearly into that of a figure 8. In some, if not all, species of Phyllochaetopterus, 

 the tubes are branched, and may be interconnected in such a way as to form 

 extensive creeping systems, the tubes in, e.g., P. socialis forming dense masses, 



