50 



developed recurrent rod (Fig. 14), which may even sometimes reach 

 the posterior end of the body. It was, however, never found to unite 

 with the body rod, so that no basket structure is formed. The rods of 

 the postoral and posterodorsal arms are simple, slightly thorny. Only 

 scattered red pigment cells are found, no accumulation of pigment in 

 the point of the arms. 



In the stage where the posterodorsal arms are beginning to develop, 

 which occurs already at the age of 8 days, — much earlier than in S. 

 pulcherrimus — the postoral ciUated band assumes a characteristic ap- 

 pearance (PI. IX, Fig. 2), with a median and two lateral sinuations. These 

 gradually grow deeper, and in the fully formed larva (PI. IX, Fig. 4) 

 the median sinuation forms a deep excavation, while the two lateral 

 sinuations form a pair of large epaulets (or vibratile lobes). The body 

 is very wide in the middle, while the oral part remains narrow, the dorsal 

 ciliated band thus making a deep sinuation at the level of the postoral 

 band. The preoral band has the same shape as in pulcherrimus, the frontal 

 area being very small. The postero-dorsal rod has appeared in the spe- 

 cimen figured in side view. In this specimen this rod lacks the usual 

 sidebranches at its base, but in sketches from other specimens I have 

 represented these small branches in the same form as in pulcherrimus, 

 which is then evidently the normal feature. 



The fully formed larva (PI. IX, Fig. 4) has the posterior end shortened 

 in the usual way through the absorption of the body rods, the hind edge 

 being nearly straight. It carries a pair of epaulets at the posterior end, 

 besides those at the base of the postoral and postero-dorsal arms, as is 

 the case also in the S. drobachiensis-lsLrya.. The arms are fairly broad 

 and flat, not very long, the postoral arms being the longest. The postoral 

 area is very much concave, the frontal area is very small. There are only 

 scattered pigment cells. There appears to be no posterior transverse 

 rod; in the said figure, it is true, a smaU rod is represented lying in the 

 place usually occupied by the posterior transverse rod. It can, however, 

 hardly be doubted that it is only a part of the body rods broken of! in 

 the course of their absorption ; its irregular shape would seem to preclude 

 the idea that it could be a true posterior transverse rod. Of course, however, 

 researches on living material will be required for definitely settling this 

 important point. 



The figure of the fuUy formed larva is partially reconstructed. No 

 doubt the outhne of the living larva wiU prove somewhat different, especi- 

 ally the body will probably be somewhat broader. Still, as the specimen 

 from which the figure was drawn is fairly well preserved, I have thought 



