,3 2 Frederick Chupm a n : 



has tigured many illustrations of these latter in refutation of some- 

 fossili figured as *' fucoids," with good reason. 



Relation alii J) of Trachydtrma to modern forms. — One of the chief" 

 determinative characters of these fossil forms in tlieir genetic rela- 

 tionship would probably be tlie morphology and arrangement of the 

 prostomial gills. These, in the Trachydtrnme now described con- 

 sist of fairly broad, unilateral, frondescent processes, haTing a 

 sigmoidal curvature and a well-defined axis. Judging by the 

 appearance of one finely preserved specimen in which the processes 

 lie back to back, they were probably paired. The nearest types of 

 ChaetopO'ds of this character are grouped in the sub-order Sabelli- 

 formia.i In these forms the branchiae are all more or less dis- 

 tinctly plumed or furnislied with secondary filaments, unlike those- 

 of the sub-Older Terebelliformia, which have simple filose or arbor- 

 escent processes. The structure of the gills in Trachyderma shows^ 

 many close points of resemblance to Dasychone, as in the flattened 

 axis, the secondary pinnules on the inner, concave side of the stem, 

 and especially in the presence of numerous eye-spots and processes 

 known as dorsal appendices. 2 These eyes have been detected on 

 several specimens, so that it is not due to any misinterpretation of 

 the surfaces of the matrix. 



The Sahellidae form their tubes of mud or sand, or of both, and 

 are usually foun<l in low water as well as to some considerable- 

 depths. In the absence of further morphological characters it is 

 advisable to place the Silurian fossil form in a new family, the- 

 T r achy derm i d ae . 



In comparing the recent worms my attention was first drawn to- 

 some of the worms of the sub-order Terebelliformia, which also* 

 make their tubes of mud or sand. In Amphitrlte johnstoni, for 

 example, '' the gills consist of a curved stem from the convex side- 

 of which arise a number of branches, themselves dichotomously 

 divided, the final branches being long, ''3 The pectinate secondar\^ 

 filaments in Trachyderma, however, are normally on the concave- 

 side of the stem, but occasionally on the outer side when the axis 

 is reflexed. The structure of the axis also agrees more closely with 

 the sabellids, and the vestiges of eyes and dorsal appendices in the- 

 fossilsi are essentially like those of this group. 



1 SeeCaiub. Naf. Tlist,., vol. ii., 189fi. Pol.v<haet Worms. Rt-nhaiu, p. 3.36. 

 •2 Op. cit.. p. £t?.7. fiiT- «iB--ft. AUo cf. DuKifchone copensl$. Krr^.^ChaJI. Znol., vol. xii , 1885. 

 Annelida I'olyehaeta. Mcintosh, p. 506, i»l. liv., fig. i. 



3 Cauih. Nat. Hist., toin. (it., pp. 3-28, 3-29 (fi<,'. 17<!rt). 



