264 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 15 



An interramal cirrus (fig. 3) is present as a long simple lobe, from 

 the second last thoracic or not before the first abdominal segment. It is 

 continued back through a long abdominal region. A subpodial lobe, 

 resembling a ventral cirrus, is present on the same segments (fig. 3). 



The modified spines of posterior thoracic segments are sagittate (fig. 

 4) and number 4 to 6 in an anterior transverse row. The uppermost 

 one projects from the neuropodium and is more conspicuous than the 

 others. These spines are present in setigerous segments 11 to 16 or 18. 

 The side of the spine directed forward is deeply grooved ; the back side 

 is smooth. The same spine, seen in different views, can thus be seen as 

 sagittate to acicular, depending on the orientation of the observer. 



Phylo felix Kinberg (1866) from Brazil has branchiae first present 

 from segment 5; the thorax consists of 19 setigerous segments; inter- 

 ramal cirri are long; modified spines are present in segments 11 to 19; 

 neuropodial fringe numbers 5 to 10 lobes at maximum development; 

 ventral fringe is present on segments 13 to 20; furcate setae have not 

 been described. 



Phylo felix, as Aricia michaelseni Ehlers (1897) from Patagonia, 

 differs from the type in that the thorax consists of 17 to 19 setigerous 

 segments; modified spines are present only in setigerous segments 15 to 

 17 or 19 (or 3 to 5 segments) ; neuropodial lobes number up to 15 

 lobes at greatest development. The details of parapodia and setae seem 

 to be identical for the individuals from Brazil and Patagonia. 



Phylo michaelseni, sensu Monro (1930, pp. 144-145, fig. 54), from 

 the East Falkland Islands, is said to have branchiae from the sixth seti- 

 gerous segment; the thorax consists of 22 segments; modified spines are 

 present in segments 14 to 22; neuropodial lobes number to 12 in a series; 

 ventral fringe is present on segments 14 to 23 with up to 15 lobes in a 

 row; furcate setae are present. This differs from P. felix Kinberg in 

 its later presence of branchiae, higher thoracic count and perhaps some 

 other characters not yet defined. 



Phylo michaelseni, sensu Okuda (1937, pp. 101-102, figs. 3 a-e), 

 from Japan, differs from the type in that interramal cirri are rudi- 

 mentary; modified spines are present in only 5 segments; neuropodial 

 lobes number to 12 in a row; ventral fringe is present on setigerous 

 segments 12 to 21 with up to 23 lobes in a row; furcate setae are present. 



Most of these records are in general agreement. The record from 

 Japan departs most widely with respect to its rudimentary interramal 

 cirri ; it may refer to a distinct subspecific or lesser systematic category. 



