ONUPHIS SOCIA. 287 



edge in each somite. The pygidium is a small, subconical body; the cu-ri are 

 broken off. 



The parapodia of the first three metastomial somites are, as in other species 

 of this particular group, attached near the ventral level. They are moder- 

 ately prominent and usually project forwards, though there seems to be some 

 variation in this regard. The first parapodia are attached at the anterior 

 border of their somite, the second and third farther caudad. The fourth para- 

 podia are attached at the middle of the length of their somite, and they project 

 more nearly directly outward. Beginning with this jjair the parapodia begin 

 to shift farther dorsad and at the seventh or eighth pair have reached near the 

 dorsal level. At the same time they have become decidedly shortened and now 

 project dorsad or a little ectad of dorsad. The fu'st parapodia are moderately 

 stout. Each bears a notocirrus, attached at its base above, which is proximaUy 

 swollen out above a somewhat constricted base and is then uniformly tapered 

 to an acute tip and is rather long, reaching, when the tip is unworn, to the 

 anterior edge of the peristomium and clearly more than halfway to the mid- 

 dorsal line. The neurocirrus is attached nearly opposite to the notocirrus and 

 is similar in form, though shorter. The setigerous process is bluntly rounded, 

 subhemispherical ; it presents a low presetal lip and a long, cirriform, distally 

 tapering postsetal process nearly as long as the neiirocirrus, but proximally 

 more slender. In succeeding parapodia, apparently to the last, the notocirrus 

 continues, but becomes shorter and very much more slender and filamentous. 

 The neurocirri on the second and third parapodia are similar in form to those 

 on the first, but those of the third pair are much smaller. (Plate 47, fig. 7). On 

 the fourth parapodia the neurocirri have abi*uptly changed to a mere flattened 

 scale and are here and on subsequent somites, therefore, essentially absent as 

 any distinct process. (Plate 47, fig. 8) . The postsetal process gradually decreases 

 in size caudad, becoming finally merely a slight conical process and not at all 

 evident as a distinct elevation caudad of the twelfth parapodia. 



The first branchiae occur uniformly on the seventeenth parapodia {i.e., 

 somite XVIII), though in one specimen the branchia appeared only on one side 

 (the right). The branchiae seem not to occur in most specimens examined 

 beyond the fifty fifth parapodia and in one ceased at the forty fifth, while in 

 another they were traced to the sixty first. Thus, while the point of beginning 

 is fixed, that of ending is variable. The first branchia, are usually simple, 

 but in fewer cases are bifilamentous or even trifilamentous on one side. (Plate 

 47, fig. 9). The second branchia in one specimen is undivided on the right 



