182 U.S. NATIONAL MUSEUM BULLETIN 23 6 



gracilis Brady, 1910, as separate species. I have the impression, 

 however, that a thorough study of specimens from the localities from 

 which the above-named species have been recorded wiU show that 

 some of these species at least must be sunk into the synonymy of 

 P. sjyinosa (Fischer) in the very wide sense defined by Lang, but such 

 a study might show as well that a certain combination of characters 

 can be restricted geographically and could be used for a badly needed 

 subdivision of this remarkable species. That Lang's views are not 

 accepted generally is demonstrated by Roe's records of P. sjnnosa 

 var. littoralis and P. spinosa var. penidllata from the Irish coasts. 



I failed to make a satisfactory division of the Ifaluk material of 

 P. spinosa and I have concluded by recording them all under that 

 name. Apart from Lang's final solution of the complicated problem 

 of variabihty in this form, he is certainly right in drawing attention 

 to the futihty of characters or the combination of characters so far 

 used to discriminate between the various varieties. 



The bulk of the Ifaluk material, though greatly variable in length, 

 corresponds with the previous form typica Lang, 1934. The anten- 

 nule is 9-segmented in the female, the 2nd furcal seta slightly swoUen 

 at the base, the setae on endopodite of the antenna normally devel- 

 oped, the exopodite of leg 1 with slender appendages. 2nd endopodal 

 segment of leg 1 lengthened, 4 to 5 times as long as wide, internal 

 seta inserting near apex. There is 1 internal seta on the 3rd endop- 

 odal segment of leg 4; on leg 5 there are 8 setae on the exopodite; 

 in the male the exopodite is 3-segmented and bears 6 setae. Anal 

 plate in some specimens finely spinulose. 



A female and a male specimen from locaHty 592, to be described 

 below, represent what might formerly have been called form peni- 

 dllata Willey, 1935; a female and a male from locality 589 approach 

 P. hornelli in several respects and will also be described below. 



In conclusion it may be remarked that the setal formula in this 

 species is apparently variable to a much greater degree than in other 

 genera of Harpacticoida and thus has only a restricted value. 



Description. — AU specimens fragmented very easily during dis- 

 section and I failed to obtain complete dissections of a single speci- 

 men. None of the specimens can therefore be completely described. 



Adult female (loc. 592), total length 0.52 mm.; proportional length 

 of cephalo thorax and abdomen is 18:17. 



Shape of body as in form typica, further characterized by relatively 

 small rostrum and fairly long, 8-segmented antennules (fig. 67). 

 Antenna (fig. G8d) with 2 well-developed spines with pectinate bristles. 

 Maxillipedes with coronula of spinules on basis. Exopodite of leg 1 

 with short, dagger-shaped appendages. 2nd segment of endopodite 

 of leg 1 styHform, 5 times as long as wide, appendages on 3rd endop- 

 odal segment of pectinate type. Legs 2 to 4 with strongly spinulose 



