On the Stmcture and Affinities of Mnestra parasites Krohn etc. 53 



With the aid of such a table as this it is verv easy to invent 

 an artificial Classification based npon one or two eharacters, but it 

 is difficult to avoid violating what appear to be very dose ties in 

 respect of other equally characteristic stmctures. 



In the system of Haeckel. a primary division of the gronp 

 was made according as to whether the radial eanals are simple or 

 forked. and as to whether the mouth is provided vrith oral tentacles 

 Mundgriffe] or not. The result was that a form Cfenaria, with 

 semipinnate tentacles was separated from the other forms with semi- 

 pinnate tentacles and was associated Avith forms with branching 

 tentacles, and vice versa Eleutken'a was classed with the forms 

 with semipinnate tentacles. But the chief objection whieh it is pos- 

 sible to raise against the Haeckelian system is that it depended 

 primarily upon the faet of whether the radial eanals fork or not. 

 and therefore upon a character which recent Observation? on the 

 variability of medusae have shown to be very variable among the 

 individuals of a species. ^uch a variable characteristic should 

 therefore not be chosen as the criterion of subfamOies. 



In 1SS7 Haetlaub having reinvestigate d the anatomy of £7^//- 

 tJferia proposed another Classification of the Cladouemidae npon a 

 new basis — upon the presence or absence of a cavity of a peculiar 

 nature used as a brood Chamber between the stomach and the exum- 

 brella. This cavity had long been known to occur in cenain genera 

 and was referred to by Haeckel as the "Scheitelhöhle" or "Kuppel- 

 höhle" or "Cavitas tholaris" and was considered by bim to be 

 homologous with that embryonic canal which originally connected 

 the growing medusa with its hydroid parent. Hartlaub contends 

 that in Eieutlieria the brood-chamber or "Scheitelhöhle" is in no way 

 homologous with the "Stielkanal" of the young medusa but is a 

 special Organ liued with a peculiar epithelium. derived from the 

 ectoderm. whieh serves the purpose of a marsupium and does not 

 communicate with the gastric cavity bnt with the snbumbreUa ca- 

 vity. Thus far the results of Haetlaub are of the highest interest, 

 but they do not justify his snbsequent procedure in attemi»ting to 

 esrablish a Classification of the Cladouemidae in whieh it is assumed 

 without reinvestigation that the "Scheitelhöhlen" of Picromma, Cfe- 

 ìta/'ia and Dendronema. all genera described and tigured by Haeckel. 

 are homologons with the very remarkable brood cavity oi Eleutheria. 

 For this assumption there is no evidence. except that Haeckel is 

 stated to have misuuderstood the nature oi the brood ca^itv of 



