596 PROCEEDINGS OF THE NATIONAL MUSEUM. 



ply the gills. The flattened trunk, together with the posterior 

 processes and egg strings, hangs free in the gill cavity. It is nearly 

 always plentifully covered with algae and various protozoa, which 

 trail backward over the egg strings and posterior processes, bind- 

 ing the whole securely together. Both the neck and the head are 

 inclosed in a very tough white membranous sheath which fits snugly 

 and aids in holding the creature securely in place. 



PAEON VERSICOLOR, new species. 



Plate 57, figs. 62, 65, 67 ; plate 58 ; plate 59, figs. 75, 76. 



Host and record of specimens. — Eight females, one with an at- 

 tached male, Avere found on the inside of the mouth of the smooth 

 hound, Mustelus cam's, at Beaufort, North Carolina, in August, 

 1905. One female has been made the type of the new species with 

 Cat. Xo. 49T62, U.S.N.M. The other seven become paratypes, with 

 Cat. No. 49763, U.S.N.M. Two other females were taken a week 

 later from the same host and have been given Cat. No. 49764, 

 U.S.N.M. 



Specific characters of female. — Cephalothorax somewhat obcordate 

 in dorsal outline, the apex being formed by the two anterior proc- 

 esses, the lobes by the posterolateral processes, while between these 

 lobes the posterior margin is deeply invaginated. The two anterior 

 processes are relatively larger than in ferox.^ are directed diagonally 

 forward and downward, and are not produced ventrally. Their flat- 

 tened inner surfaces entirely cover the proboscis and mouth parts. 

 A pair of lateral processes project from the sides of the cephalo- 

 thorax behind this frontal pair, and the bases of the two pairs are 

 fused. The dorsal portion of the head is greatly inflated and sends 

 out on either side diagonally backward and upward a large spherical 

 process. The two do not quite meet on the midline dorsally, leaving 

 a deep sinus between them. The ventral processes, instead of being 

 diagonal as in ferox., are at right angles to the head axis and are 

 fused across the midline, so that they appear as one. On the ventral 

 surface of these processes is found on either side another jointed, 

 finger-like proturbance similar to those in ferox. It, too, is solid, 

 contains what appear to be muscles, and is split at the tip. It is 

 on the postero-ventral instead of the postero-dorsal surface of the 

 process, but this difference in position is probably due to the differ- 

 ences in the processes themselves. Whether these protuberances are 

 really migrated appendages can only be determined by following 

 their development, but they certainly resemble appendages. In 

 Sphyrion the maxillipeds remain close to the midline, but are almost 

 concealed by a pair of fused ventral processes just in front of them, 

 and we might suppose that here similar processes by an increased 



