NO. 1573. PARASITIC COPEPODS— WILSON. 463 



Neither these nor the ones on the bases of the swimniini!; k'<;s seem 

 as well formed and efficient as those of the preceding subfamily, 

 and they are evidently not of much service in prehension. 



The attachment of the parasite to its host being thus accomplished 

 entirely by means of stout claws buried in the fish's flesh, it follows 

 that there can be no loosening of the hold after the death of the lish. 

 These parasites certainly never drop ofi' into the boat or anywhere 

 else, as was claimed for the genera in the preceding subfamily. Both 

 the antennae and the maxillipeds are set so firmly in position that 

 they can not be loosened without breaking them. And even after 

 the death of the parasite it is usually necessary to cut around these 

 appendages if one would remove the animal entire. 



On being placed in an acpiarium they are more helpless than the 

 Pandarina?, and l)oth sexes can only lie in the place where they are 

 put, with an occasional weak and spasmodic movement of the swim- 

 ming legs. Hence they can be kept alive only a very short time, 

 shorter even than in the case of the Pandarina^. 



Although they resemble the preceding subfannly in that the cara- 

 pace is not arched and converted into a disk to retain moisture after 

 the skin of the host may have dried, yet they are kept alive by 

 another means for quite a period. It has already been noted that 

 they are gregarious and gather together in colonies, whose combined 

 lacerations form a deep pit extending through the skin and into the 

 flesh of the host. Such a pit does not dry as quickly as other por- 

 tions of the surface of the host's body, and the parasites, being fas- 

 tened in its very bottom, are not only protected from abrasion, but 

 are also kept moist long after the fish's skin has become dry and 

 parched. 



In this way they may sometimes be fountl alive on sunfish brought 



in by the fishermen. 



MORPHOLOGY. 



The chief dift'erences in the morphological details of the present 

 subfamily are such as result from a further degeneration in conse- 

 quence of the fixed position of both sexes, and they are clearly brought 

 out in the diagnosis which follows. 



ONTOGENY. 



The development of this subfamily has never been observed, but 

 it must be similar to that of the Pandarina"., if we may judge from 

 the size, shape, and coloration of the eggs and from the chalimus of 

 OrtJiagoriscicola, described on page 476. 



We find in the latter a pair of broad and ribbon-like frontal fila- 

 ments exactly like those on the chalimus of Perissopus. In this case, 

 also, it was the male which was thus found, indicating that the two 

 sexes mature at different periods. 



