Vol. 50.] SYSTEMATIC POSITION" OF THE TRILOBITKS. 



427 



It is only when the limb is flattened out under a cover-glass that 

 the exopodite assumes its true morphological position as a dorsal 

 appendage of the endopodite, branching off laterally in the transverse 

 plane. Further, the flat rowing exopodite of Apus is supplied with 

 a fringe of sensory hairs. These hairs are very marked on the 

 exopodite of Triarthrus, which, as above noted, has the same position 

 with reference to the endopodite as in Apus. 



Proximally to these two branches, Ajjus has a gill on the dorsal 

 side of the limb. This organ is either not uncovered in any of the 

 described specimens of Triarthrus, or else was quite rudimentary 

 in these animals. But Walcott's researches have led him to the 

 conclusion that the trilobites possessed gills in the typical place, 

 and often, in adaptation no doubt to their manner of life, highly 

 specialized structures. 



Fig. 13. — Section o/^Calymene 

 senaria. (After Walcott.) 



,nj0M 



Fig. 14. — Corresponding section 

 through Apus (Lepidurus) 

 spitzbergensis, Bernard. 



" So far, then, we have the limbs of the trilobites fundamentally 

 of the same type as those of Apus. But the question of prime 

 importance still remains to be answered — were the trilobite-legs 

 phyllopodan, or, considering their more filamentous distal portions, 

 do they show any traces of having been originally membranous 

 appendages with broad transverse insertions ? 



Walcott's figures appear to me to leave no doubt on this point. 

 The sections (figs. 13 & 15, from Walcott) are almost exactly 

 paralleled by longitudinal sections of Apus (figs. 14 & 16),- so far, 

 that is, as the section through the limbs is concerned. The limb of 

 Apus has a long transverse attachment, partly to the ventral and 

 partly to the lateral surface of the body. Sagittal sections cut 

 laterally (fig. 16, p. 428) show the divisions between the limbs running 

 high up the sides of the body as in the corresponding section of 

 Calymene senaria (fig. 15, p. 428). Fig. 13 shows a section through 

 the same trilobite, and fig. 14 one through Apus, passing through the 

 lobate basal portion of the limbs farther in, that is, nearer to the 

 median plane. On comparing the four sections here given, we thus 

 have, in both animals, the attachments of the limbs occurring not 

 only in tangential sections, but in those taken much farther iu 

 towards the median line. This can have but one explanation, namely, 



