304 M. T. THOMPSON. 



almost adult form but lack abdominal appendages. The sexu- 

 ally mature P. cxocccti evidently possesses abdominal appendages 

 that are subdivided into several branches. The relative develop- 

 ment of abdominal appendages compared with the "head" in 

 the oldest specimen of the P. filosa series - - " F ' - suggests the 

 possibility of a somewhat similar phenomenon here, to the extent 

 that individuals may occur with the general appearance of an 

 adult, but with less complexly cleft appendages. Such ap- 

 proaches in development to an adult bodily form before the 

 abdominal structures have fully shaped themselves deserve notice, 

 if only from the standpoint of the systematist. They may readily 

 become not unimportant sources of confusion. 



Moreover with P. varians the abdominal appendages are vari- 

 able even in the adult animal, Steenstrup and Liitken recording 

 them as bifid for several mature individuals (p. 411) although 

 they are typically uniramous. With P. filosa the abdominal 

 appendages of all the specimens of mature form that I have had 

 opportunity to examine, were complexly subdivided into fila- 

 mentous branches. The occasional presence of one or more 

 simple or few-parted appendages at the anterior end of the series 

 was the only variation found. This occurred in two out of 

 seventy-nine specimens from Orthagoriscus niola in a collection 

 loaned me by the Museum of Comparative Zoology. At least one 

 of these was sexually mature and bearing egg-sacs and the same 

 was true for a single specimen out of fifteen examples from the 

 swordfish, which individual had three of the anteriormost ab- 

 dominal appendages tri-ramous only. The number of the ab- 

 dominal appendages is fairly constant in Pcndla filosa ; varying 

 from 21 to 30 pairs. Twenty-two or twenty-five pairs are the 

 commonest. Of my larvse : "A" had 22 pairs, " B " 21, "C 

 23, " D " 21, " E" 20 and " F" 22. 



On comparing the youngest Pcndla larva known, the imma- 

 ture P. varians figured by Steenstrup and Liitken, with that stage 

 of the ontogeny of Lcrn<za brancliialis at which fertilization and 

 attachment to the final host takes place - -the " begattungssta- 

 dium " of Claus' ('68) description -- or to the corresponding 

 phases of other lernaeoid copepoda, the resemblance in general 

 structure is striking. And it appears probable, therefore, that 



