302 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 7 



Stage of development. In others such is not the case ; a few more precocious 

 individuals soon attack the more retarded ones, so that the cellular con- 

 tents are withdrawn (see below). 



The earliest embryos in which structures are visible have a slightly 

 lobed appearance (pi. 47, figs. 30, 31). The oral area of the anterior 

 region is well ciliated. There are weak indications of segmentation and 

 shallow transverse grooves on the ventral side, set apart by rows of short 

 cilia. No long trochal cilia are observable, such as characterize typical 

 planktonic polytrochs. The region of the alimentary tract is filled with 

 yellow granular yolk particles. Not long thereafter prototroch and telo- 

 troch are present, and the cephalic region is better defined (pi. 47, 

 fig. 32). This short, rotund form gives rise to a triangular-shaped, 3- 

 segmented larva (pi. 47, fig. 37). The prostomium is clearly marked with 

 2 pairs of dark eyespots dorsolaterally and a pair of clear pouches opening 

 anteriorly. On the ventral side the oral aperture is well ciliated, the lower 

 lip bounded by large columnar cells. Before setae have emerged, elongate 

 cells in the setigerous sacs may be seen with slender, rodlike spines. Setae 

 that finally emerge are entirely slender, capillary, provisional, without 

 limbate region such as those in the adult. These setae from the first 3 

 segments elongate rapidly and extend posteriorly so as to surpass the anal 

 end. The alimentary tract is complete when the full length of the setae 

 has been attained. If a cocoon be broken open in which such larvae are 

 encased, they squirm rapidly out of the case and swim about, alternately 

 holding the setae along the sides of the body or thrusting them laterally 

 (pi. 47, fig. 37). 



Several such cocoons were isolated in separate culture dishes, on 

 March 2, to determine the fate of the larvae if left undisturbed, and 

 whether they could effect their own release from the capsule. After 6 

 days they were still moving about in the cocoon, though considerably im- 

 peded by the proximity of others in the case. Many setae had been pulled 

 out and were packed in the stalk end of the capsule. After 8 days the 

 larvae were dead. Rate of survival of other cocoons differed somewhat, 

 but in no case did they bring about their own release from the cocoon. It 

 seems that an external stimulus must be applied, such as might be pro- 

 duced by the brooding individual in passing back and forth through the 

 tube. 



Other similar cocoons were broken open, peniiitting the young to 

 escape. These were kept in culture for several weeks. They were capable 

 of foraging for themselves and kept alive with no difficulty. When sup- 



