NO. 3 HARTMAN : POLYCHAETOUS ANNELIDS 325 



were developed from the first segment. Johansson proposed that they 

 develop from the second last prothoracal segment, but that, preceding 

 this, there are several (from 3 to 5) segments that have come to be more 

 or less reduced. They are suppressed by the unusually great development 

 of the paleal segment (designated segment 2 by Johansson) and have 

 fused with one another and with the prostomium so that all indications 

 of septa are efifaced. The segment most clearly discernible is the one pre- 

 ceding segment 2 (designated segment 3 by Johansson), since here the 

 setae, though reduced and embedded in the opercular musculature, are 

 still present. This is the condition in species of Sabellaria; in those of 

 Idanthyrsus and others where nuchal hooks are present, this segment 3 

 is believed to be represented by the nuchal hooks. Segment 4 (preceding 

 segment 3) is so reduced that only the presence of spinal nerves indicates 

 a former existence. It is likely that a fifth segment preceded it, indicated 

 by large nerves in the musculature on the ventral side of the opercular 

 paleae. In conclusion, therefore, Johansson believes that the opercular 

 paleae originate from the notopodium not of a single segment, as Meyer 

 thought, but of parts of 3, 4, or even 5 segments. 



In addition, Johansson (1927, p. 25) believes (and I concur in this 

 view) that in all members of the Sabellariidae there are fundamentally 

 2 rows of opercular paleae. Even in Sabellaria (p. 337), where 3 rows 

 of paleae are usually considered, there are 2 rows so arranged that those 

 of the inner and middle rows belong to a single series, alternately directed 

 mward and outward. This view is strengthened by the fact that the 

 paleae of the outer row often approximately equal in number the sum of 

 the middle and inner rows. In addition to these stout, visible paleae, fine 

 rudimentary capillary setae are present (in Sabellaria) embedded in the 

 thick musculature on the dorsal side of the operculum. They are signifi- 

 cant, since they herald the presence of another segment, and are the 

 homologues of the heavy nuchal hooks in some other genera {Idanthyrsus 

 and Lygdamis) . 



It might be suggested that segment 3 (preceding segment 2 and so 

 designated by Johansson) actually succeeds it posteriorly, since it is repre- 

 sented by a single ramus, the notopodium, and that the neuropodium of 

 the first anterior thoracic segment is its corresponding ventral part. This 

 would reverse the order of segments 2 and 3 (of Johansson), but would 

 in no way change his conclusions. 



In conclusion, therefore, it might be said for the segments which 

 manifest themselves in the adult that the opercular crown, including 

 the paleal spines (whether in 2 rows or 3 apparent ones), represent the 



