652 MR. R. T. cu'^NTUER OX [June 18, 



attached to the outer skeleton of some actively locomotive animal, 

 svich for instance as one of the reptiles which abounded in 

 Mesozoic times, and even at a later date, then, on the further 

 supposition that its host sometimes made excursions from the sea 

 into fresh-water, we should have a means by which the hydroid 

 might be introduced." 



Unfortunately for this theory, there is no evidence of any 

 fixed hydroid stage in Limnocnida, indeed the evidence is all 

 against the existence of one : Aurelia aurita, Cramhessa tagi, 

 and Mceotias have all left the sea without the aid of a marine 

 reptile : Halmonises lacustris was probably helped into fresh-water 

 by tidal agencies alone, and, as Sir Ray Lankester has suggested, 

 Limnocodium may not pass through a fixed hydroid stage at all ! 



The Systematic Position of Limnocnida. 



The position of Limnocnida in the Haeckelian System is 

 still a matter under discussion. Judged by the chief diagnostic 

 characters of endodermal sense-organs and manubrial gonads, 

 Limnocnida should be one of the Narcomedusa? ; but, on the other 

 hand, Narcomedusse are distinguished by characters so peculiarly 

 their own, that Limnocnida cannot be considered as being at all 

 closely related to them. Moreover, I know of no Anthomedusan 

 which at all resembles it, except in the position of the gonads 

 on the manubrium. 



On the other hand, the grade of development which Limnocnida 

 has reached is very closely paralleled by that of the other fresh- 

 water medusa Limnocodium, more especially in regard to the 

 tentacles and sense-organs, as I have already pointed out in an 

 earlier communication. More recently (1903) Seitaro Goto has 

 published illustrations of the sense-organs of the marine OUndioides 

 clearly showing them to be of the same type. This distinguished 

 Japanese naturalist has associated OUndioides, Halicalyx, and 

 Gonionema with the Olindiadae, which he regards as a subfamily 

 of Eucopidfe (w^e believe AATongly) and near which he believes that 

 both Limnocodium and Limnocnida may be most naturally placed. 



The Olindiadae are defined as Eucopidfe (see below) with two 

 sets of tentacles, velar and exumbrellai-, the former springing 

 close to the base of the velum, and the latter at variable distances 

 from it, but always from the exumbrella and connected with the 

 circular canal by endodermal roots : marginal vesicles numerous, 

 two on either side of the bases of the exumbrellar tentacles : 

 manubrium well developed and quadrate, with distinct lips : 

 radial canals four or six : gonads primarily continuous folds of 

 the walls of the radial canals : with an adhesive disc on each 

 exumbrellar tentacle. 



The greater part of this definition applies quite well to the fresh- 

 water Limnocodium sotoerhyi, with the exception that this species, 

 not being a creeping form, has tentacles without adhesive discs, . 

 and has also sac-like e'onads. 



