TBRBBBLLIDJB. 99 



cirriform branchige. This author therefore made a useful classification for the period, but 

 it lacked the precision and breadth which the study of other structural, features subse- 

 quently gave, and he included the Ampliaretidae in the family instead of separating them. 



Dr. Johnston (1865) made the Terebellidse his fifteenth family out of a total number 

 of sixteen, but his description includes the characters of the Amphictenidae as well as 

 those of the Terebellidse. Accordingly the presence or absence of a front armed with a 

 row of stiff bristles constituted his main divisions. The Terebellidaa were arranged in 

 three groups, the first having three pairs of branchiae, the second two pairs, and the third 

 one pair — a classification more or less on the lines of De Quatrefages. 



Leon Vaillant 1 (1865), who had studied an imperfect example of a Polycirrid, thought 

 that the tentacles indicated a series of buds. The error was noted at the time, and Ch. 

 Gravier 2 has since fully explained the misinterpretation. 



Claparede (1868) divided this family into two tribes, viz., Branchiate and Abranchiate, 

 and he made special remarks about the tori imcinigeri, which he ranged under five heads 

 according as their hooks pointed forward (" rangee progressive"), backward ("r. retro- 

 gressive"), alternated ("r. alterne"), similar to the last but wider apart ("r. engrenantes"), 

 and lastly an anterior and posterior row with opposite points connected by a curved series 

 at one end (" r. parabolique "). He found young forms at Naples easily discriminated by 

 this method. He also alludes to the tendinous processes attached to the hooks under the 

 title " soies de sontien " or " soies-tendons " in the posterior region. In his later publi- 

 cation (1870) the author corrects A. Costa and others, who insisted on the presence of 

 vessels in the tentacles of this family. Only the coelomic fluid penetrates into these organs. 



In Claparede's posthumous work on the structure of the sedentary Annelids (1873) 

 many interesting remarks occur on the Terebellids, especially in connection with Terehella 

 flexuosa, a few of which may be summarised. He lays great stress on the supporting 

 functions of his clypeal tissue in Terebellids. It not only furnishes the external shield- 

 like thickenings in the mid-ventral region, but passes internally and forms a mass in the 

 perivisceral chamber supporting the nerve-trunk and vessels. The connective tissue, 

 again, is amorphous, with nuclei. The new hooks are formed at one extremity of the 

 torus only, and the torus is hypodermic. He states that the dorsal longitudinal muscles 

 are united in the Terebellids and that there is a lateral dorsal muscle. The circular 

 muscle forms only an isolated cordon in each segment in the thoracic region, and by this 

 he refers to the lateral region below the bristles ; yet in his figure (Plate IX, fig. 7) it is 

 seen that he really means the circular or transverse fibres outside the nerve-cord. He 

 considers the oblique muscles as of minor importance in the Terebellids, yet they are 

 largely developed in sedentary annelids. The nuclei of the muscular fibres are external 

 as in Oligochaats. 



The coelom is divided into two chambers by a longitudinal median mesentery 

 enclosing the dorsal and the ventral vessels. Moreover, no communication exists between 

 the perivisceral chamber in Polychasts and the exterior as in Lumbricus. A large part 

 of the body in Terebellids is devoid of dissepiments. In certain forms a diaphragm 



1 ' Ann. Sc. nat./ t. iii, pp. 242 — 250, pi. hi. 



2 ' Ball. Soc. Philomath. Paris/ 1906, pp. 1—14, with text-figs. 



