I'.M; TACE SKO(iSHKR(i 



No males ol thr sub-gcucru Munopni ;iiul i^vp/ionustru ure known so iiir; on account oi 

 this these two units could not be indudoil in the above table. 

 rros9vphorus. As is seen from this table the genus Crossophorus is decidedly opposed to all the otlici 



units included in the table, as its male first antenna is quite without suctorial organs on the 

 b- and o-bristles. ^^^lether this absence of suckers is primary or not is, of course, a question 

 impossible to decide with certainty at present; it seems, however, fairly probable to me that 

 it is primary. \\'ith regard to the endopodite of the male second antenna this genus is primitive. 

 In a number of other characters as well Crossophorus seems to have retained a certain primitive- 

 ness, for instance with regard to the endites of the mandible (cf. p. 171 above). It seems 

 to me rather probable that this genus is to be considered as being in several respects 

 the most primitive one in this sub-family. 



A close comparison fully confirms the fact that there is a contrast between 

 Crotysophoriis and the other genera belonging to this sub-family, ahnost all the organs in 

 the former genus showing a more or less divergent type, for instance first and second antennae, 

 the mandible, the seventh limb, and the furca. As a matter of fact there can be no doubt 

 that this genus occupies a rather isolated position in tliis sub-family. It might be most 

 convenient to set this genus up as a representative of a special group within this 

 sub-family, in contrast systematically with all the other genera, or perhaps even to 

 distinguish it as a sub-family, CrossopJwrinae, i. e. as a group equivalent systematically 

 to Cypridininae and Philomedinae. 



For the relation of the genus Crossophorus to the last-mentioned sub-family see above, p. 178. 

 The other units. All the other iinits of the sub-family Cypridininae mentioned above seem to me to be 



the result of a not very extensive variation in different directions of one and the same funda- 

 mental type. 



According to the table given above these units may be divided into three group as follows: 

 I. Gigantocypris 

 11. Codonocera 

 111. Doloria, Vargvla, Macrocypridina, Cypridina (s. str. et meo) antl Cypridinodes. 

 This classification, although based exclusively on the equipment of the b- and c-bristles 

 on the male first antenna, seems also to represent the mutual phyletic position of 

 these units. For a close investigation of the general organization of these forms seems 

 to give the result, partly that Gigantocypris and Codonocera are mutually rather 

 different, partly that each of these two genera presents a certain contrast to the five sub- 

 genera included in the third group. Whether Gigantocypris or Codonocera is to be considered 

 as more closely related to the five sub-genera in question seems to be difficult to decide at 

 present. Probably, however, Gigantocypris represents the type that differs most from these. 

 The sub-genera Doloria, Vargida, Macrocypridina, Cypridina and Cypridinodes are 

 certainly closely related to one another. They are also closely related to Monopia 

 and Siphonostra, the two sub-genera that are not included in the table given above. To 

 show conclusively their natual mutual relations is an exceedingly difficult — if not quite 

 impossible — task. I only wish to make the following statements: 



