Stiulir'S im iiiiiriin^ Ostrarods 36( 



Bvgning soin lios de 2 foregaaende Slaegter; men de er forholdsvis inindip kraftigt udviklede, 

 og de til 8v0mmegrenen faestede Borster er ualmindelig korte, med C'ilieringen grovere og 

 mindre taet. I Virkeligheden kan disse Lemmer lios Hunnen kun uegentlig kaldes Svemmc- 

 antenner. da de ikke benyttes til Svomning, men kun som et Slags Arme .... ved Dyrets 

 langsomme krybende Vevaegelser paa Havbunden."* The drawing with which Gr. 0. Sars 

 illustrates this description shows an antenna of about the same structure as the larval one, 

 i. 0. with relativelv short, unbroken bristles, well pointed distally, on the exopodite; curiously 

 enough in this drawing all the bristles of the exopodite have natatory hairs. From this 

 description and figure it seemed to me probable that the peculiarity of breaking-off the 

 natatory bristles would also be a characteristic of this species. In order to be absolutely 

 certain on this point I wrote to Professors Gr. 0. Sars and G. S. Brady asking for permission 

 to investigate their specimens. Both these investigators were kind enough to send me several 

 specimens. Among the specimens sent by Professor Sars there was only one (probably) mature 

 female; this specimen unfortunately, however, was represented only by two empty valves. 

 Among Professor Brady's specimens there was a complete female with very large eggs in the 

 brood chamber. Contrary to G. O. Sars's statement this specimen had, on the exopodite of 

 the second antenna, like Ph. ( Sderoconcha) Affellofi, relatively short and quite bare bristles 

 on the second to the fourth joints; the bristles on the following joints were long natatory 

 bristles of the same type as in the female of Ph. (Ph.) globosa during its pelagian stage. 



What is the connection between these facts and the view put forward by G. W. MtlLLER, 

 1908, that the peculiarity of breaking off the natatory bristles in the genus Philomede.s is not 

 a phenomenon of convergence, but that it is to be referred to a common inheritance? 



It is obviously tlifficult to fit them in with this theory. Pliilomedes ( Sderoconcha) Afpelloji 'i'h<t hrvaking ofj »/ 

 and Ph. (Ph.) rotunda represent two types rather strongly differentiated from Ph. (Ph.) globosa "'" "f'^i^'-yhrisac. 

 and Ph. (Ph.) assimilis; the first-mentioned species especially differs comparatively greatly 

 from the others. In all these four forms the natatory bristles are broken off. In Ph. (Ph.) 

 Eugeniae, which is certainly very closely related to Ph. (Ph.) globosa and Ph. (Ph.) assimilis. 

 and in Ph. (Scl.) Folini, which is very closely related to Ph. (Scl.) Affellofi the natatory bristles 

 remain unbroken throughout the whole life. 



Contrary to G. W. MUller's view, it seems to me necessary to assume that the peculiar 

 character of breaking off the natatory bristles of the exopodite of the second antenna in tiic 

 genus Philomedes is not tlie result of common inheritance but of convergence. 



It is of course impossible at the present time to give any certain causes for this pheiio- 

 monon. The following facts are, however, striking: 



1) All the species (five) of this genus from warm or temperate seas that were in- 

 vestigated with regard to this character proved to have long, unbroken natatory bristles 

 during the whole year. 



* The sflconrl piiir of antoniuii' in tin- fein;ile rorlainly show, oi\ tlie whole, the satiie .sfrufliire as in llie Iwo 

 preredini; jrenera: Ijul they arc (•omparalively less powerfully (leveloped. and the bristles that are attached to the 

 natatory branch are unusually short, with their hairs coarser and situated less densely. .\s a matter of fact these 

 limbs in the female cannot really be called swimming antennae, as they are not used foi' swimminfr, but only as a sort 

 of arms . . . . iti the aiiimnrs stow clnwllni; movements at the liotlom of the sea. 



•Jiiverge/icrf 



