Tl' tack skohshhro 



tliat whuli cliariictorizt's thi' ivci'iil C y p r i d i ii i d ,s; there are. both anatoiiiieal aud lue.eliani- 

 lal reasons against this. This method of swimnung seems to presuppose the dominance of the 

 exopodite. This circumstance seems also to have been noticed by G. W. MtlLLER; this author 

 writes, 1894, p. 193, as follows:* ..DaU bei einer fast ausschlieBlichen Verwerthimg der 2. Antenne 

 als SchwimnifuB der Innenast schwindet, scheint verstandlich, denn er verdankt seine Erhaltung 

 als kleiner Rest nur der Function als Greiforgan beim (^ und diirfte diese bereits bei der gemein- 

 samen Stammform der Mi/atlocopa besessen haben." It will probably be sufficient to point 

 out in this connection that all the forms that swim in this way (all the C y p r i d i n i d s, 

 all the genera of H a 1 o c y 2> r i d s except Thaumatocypris) have the endopodite reduced; 

 this branch does not help as a natatory organ. On the other hand, in Thaumatocypris and the 

 P o 1 y c o p i d s, \vlii( h are, as we know characterized by another method of swimming, both 

 the exopodite and the endopodite are always well developed. 



It seems to me most probable that the rostral incisur swimming is a later acquisition. 

 It even seems not impossible that this method of swimming has arisen and been developed 

 independently in the two groups, C y p r i d i n i d s and H a 1 o c y p r i d s. This idea seems 

 to be decidedly supported by the fact that Thaumatocypris, the genus that is in many respects 

 the most primitive of aU the H a 1 o c y p r i d s, does not have this method of swimming, but 

 swims in quite a different way. It must, of course, be considered as very improbable — not 

 to say entirely impossible — that the Cypridinids diverged from the Halocyprids 

 after Thaumatocypris. 



Can we assume that any other of the three methods of swimming described above as 

 occurring in the recent Ostracods is primitive in this group? 



It seems to be impossible to assume that the method of swimming that characterizes 

 the genus Thaumatocypris is original; as far as I can see this method needs long processes on 

 the shell (cf. below, the chapter on adaptation to a planktonic life) and such processes could 

 scarcely have characterized the shells of the Protostracods. 



There remains consequently only the method of swimming that we found as characteristic 

 of the Polycopids and a number of the C y p r i d s. But it does not seem possible to 

 consider this either as primitive in the Ostracods, as both the position of the Poly- 

 copids and the C y p r i d s in the Ostracod system and the details in the development of 

 this mode of swimming seem to support very decidedly the idea that this mode of swimming 

 has arisen and been developed independently in these two groups. 



Is it not really at least equally probable that the ancestors of the Ostracods were 

 not freely swimming but crawling forms — although their powers of crawling were not quite so well 

 developed as in a number of recent forms, e. g. Nesideids and Cytherids? By this 

 I do not, of course, mean to state decidedly that they had a crawling life and that they lacked 

 all power of swinmaing, but I only wish to point out that this possibility does not seem to me 

 excluded. Before we have succeeded in showing quite definitely that this possibility is out of 

 the question it does not seem right to put forward an assumption that the opposite state of 

 affairs is the correct one — at least the matter should not be put in such as definite way as 



* I leave allogelher out of coiisidcratiun llir illngical di'dui linn in lliis slaliMiii'iit. 



