376 THE ANNELIDA POLYCHAETA. 



sublongitudinally wrinkled. The succeeding segments are smooth; they are 

 undivided above but show ventrally a vague transverse line marking off a shorter 

 caudal from a longer anterior part. The anus is small, circular, and terminal, 

 opening somewhat more dorsad than ventrad, the ventral border projecting 

 somewhat caudad. The setigerous segments are all distinctly separated; 

 in the type they are about one thu'teenth as long as wide and further caudad 

 may be as much as one seventh. 



The branchiae are numerous, and their precise proportionate length cannot 

 be determined because of their tangled and broken condition, but would seem 

 at most not to exceed four times the diameter of the body. The branchiae occur 

 in pairs on all the segments of the anterior region and caudad in the large speci- 

 men, as judged by the scars, to about the seventeenth segment from the caudal 

 end. In the anterior region each branchia is inserted at the edge of the dorsal 

 continuation of the elevated ridge bearing the setae ; the distance of the branchiae 

 above the notopodial tuft increases caudad to the middle region of body, over 

 which the distance remains essentially constant, this being in the type about 

 3.2 nrni. The vertical ridges above the notopodia become less distinct and finally 

 obliterated caudad. The special branchiae occm- in a tuft on each side of the 

 first setigerous somite. The exact number in each tuft could not be determined, 

 but in the paratype it seems to be between twelve and eighteen. They do not 

 extend far up the side, there being a wide naked region on the dorsum. 



The setigerous tubercles are, as usual, well separated, though in the middle 

 region the distance between them is much less than that between notopodium 

 and the branchia (1.8 mm. as against 3.2 mm.). They are borne upon a ridge, 

 the upper end of which is extended beyond and bears at its edge the branchia. 

 The ventral spines, as nearly as could be determined, appear first on the fifty 

 fifth somite in one specimen and the sixty fifth in the other. The dorsal spines 

 could be found first on about the one hundred and twentieth somite in one, and 

 the hundred and sixty fifth in the other, but the bad condition of the specimens 

 renders some error likely. The most anterior ventral spines are pale, these 

 caudad soon becoming darker and stouter. The number of ventral spines is 

 two or three, three being the maximum. These spines in the posterior region 

 are dark, stout, and straight, and show both cross and longitudinal striations. 

 (Plate 70, fig. 3). The anterior spines are paler and rather less stout. The 

 dorsal spines are paler and more slender. The capillary setae are long, those of 

 the notopodia being longer than the neuropodials. They are finely closely 

 serrate on one side throughout their length. (Plate 70, fig. 3, 4). 



