16 THE CRASPEDOTE MEDUSA OLINDIAS AND SOME OF ITS NATURAL ALLIES. 



any of the tentacles. I may, however, be allowed to suggest that the exumbrellar 

 tentacles (stiff tentacles standing out sharply at right angles to the bell and sprinkled 

 over with wart-like protuberances of a deep purjile color) will probably be found on a 

 closer examination to be provided with an adhesive disk near the tip. The mucous 

 glands in similar positions of Olindias have been overlooked. 



Haeckel says that he at first placed Olindias in the iEquorida?, but that the 

 structure of the marginal vesicles shows it to belong to the Trachomedusse. Mayer 

 ('99) also regards Gonionema as a trachomedusan ; but Murbach ('95), who has stud- 

 ied its development, says that there is true alternation of generations, the larva pass- 

 ing through a hydrula stage. Haeckel places Gonionema in his leptomedusan family 

 Cannotidse, which is characterized by the absence of marginal vesicles, the possession 

 of four- or six-branched, forked or pinnatifid radial canals, in the course of which the 

 gonads lie. On the systematic position of Halicalyx neither Fewkes nor Mayer says 

 anything. 



It is evident from a perusal of their works that Haeckel and Mayer have been led 

 to regard Olindias and Gonionema as belonging to the Trachomedusae by the struc- 

 ture of the otocysts, and the only peculiarity of these organs that could have impressed 

 them appears to me to be that the otolith is provided with a stalk, by means of which 

 it is attached to the wall of the vesicle. But the presence of a stalk taken by itself 

 seems to me to be of no systematic importance, since, in all the Leptomedusse thus far 

 described or observed by me, the otoliths are attached to the wall of the vesicle, and 

 it is simply a question of the comparative length of the intervening portion that deter- 

 mines the presence or absence of the stalk. Further, if one compares the stalk of the 

 vesicle of any of the medusae above mentioned with that of such undoubted Tracho- 

 medusse as Glossocodon or Liriope, there are some differences, which, taken in conjunc- 

 tion with the observations on the development of the otocyst, point to a fundamental 

 difference between the two. The otolithic stalk in Glossocodon and Liriope is very 

 distinctly set off both from the protoplasmic mass surrounding the otolith and the 

 wall of the vesicle; but in all the genera of Olindiadse the stalk passes on gradually 

 into the periotolithic mass on the one hand and the vesicle wall on the other. We 

 have also seen that in Olindioides the endoderm takes no part in the formation of 

 the otocyst. 



Among the differences set up by the Hertwigs ('78) for the two types of otocysts, 

 one characteristic of the Trachylinse and the other of the Leptolinse, is one of innerva- 

 tion. According to the observations of the two brothers, the otocyst is innervated 

 in the former group by the upper (outer) nerve-ring, and in the latter by the lower 

 (inner). In Olindioides and Gonionema the otocysts appear to be supplied from 



