332 LARVAL FORMS. 



Their mode of development is well illustrated by the case of Terebella 

 nebulosa (vide Milne-Edwards, No. 347). The embryo is at first a nearly 

 spherical ciliated mass. One end slightly elongates and becomes free from 

 cilia, and, acquiring dorsally two eye-spots, constitutes a prae-oral lobe. 

 The elongation continues at the opposite end, and near this is formed a 

 narrow area free from cilia. The larva now has the same characters as the 

 atrochal Eunicidan larva described above. It consists of a non-ciliated 

 prae-oral lobe, followed by a wide ciliated band, behind which is a ring-like 

 area free from cilia ; and behind this again a peri-anal patch of cilia. The 

 ring-like area free from cilia is, as in the Eunicidan larva, the region which 

 becomes segmented. It soon becomes longer, and is then divided into two 

 segments ; a third and fourth etc. non-ciliated segment becomes succes- 

 sively interposed immediately in front of the peri-anal patch ; and, after 

 a certain number of segments have become formed, there appear on some of 

 the hinder of them short tubercles, provided with single setae (the notopodia), 

 which are formed from before backwards, like the segments. 



The mouth, anus, and intestine become in the meantime clearly visible. 

 The mouth is on the posterior side of the ciliated band, and the anus in the 

 centre of the peri-anal patch. 



The ciliated band in front now becomes contracted and provided with 

 long cilia. It passes below completely in front of the mouth, and constitutes, 

 in fact, a well-marked prae-oral ring, while the cilia behind constitute an 

 equally marked peri-anal ring. The larva has in fact now acquired all the 

 characters of a true telotrochal form. 



Only a comparatively small number of Chaetopod larvae 

 remain permanently telotrochal. Of these Terebella nebulosa, 

 already cited (though not Terebella conchilega), is one ; Poly- 

 gordius, Saccocirrus and Capitella are other examples of the 

 same, though in the latter form the whole ventral surface 

 becomes ciliated. 



The majority of the originally telotrochal forms become 

 polytrochal. 



In most cases the ciliated rings or half rings of the polytro- 

 chal forms are placed at equal distances, one for each segment. 

 They are especially prominent in surface-swimming larvae, and 

 are in rare cases preserved in the adult. In some instances 

 (e.g. Nerine and Spio) the ventral half rings, instead of being 

 segmentally arranged, are somewhat irregularly distributed 

 amongst the segments, so that there does not seem to be a 

 necessary correspondence between the ciliated rings and the 

 segments. This is further shewn by the fact that the ciliated 

 rings are not precursors of the true segmentation, but are 



