l8g3 . THE LUCERNARIANS. 207 



extinct, since it cannot compete for long in the same sphere with the 

 better equipped varietal form. 



The clue that led me to this conclusion as to the phylogeny 

 of the Lucernarise was the fact of finding great and remarkable variation 

 or inconstancy in Halidystus octoradiatus in those bodies termed colleto- 

 cystophores, and which are generally supposed to be homologous with 

 the marginal organs — tentaculocysts, etc. — of the medusae. Such 

 variation consists, as I have recorded elsewhere, (a) in duplication of 

 the organ, (b) in the frequent suppression of one or more in the same 

 animal, (c) in the still more frequent reversion in structure to that of 

 the form of the normal papillae or tentacles as seen in the marginal 

 clusters of these appendages. This gives us the fixed fact of great 

 variability in the marginal bodies. Now inconstant presence and frequent 

 reversion are known to be the accompaniments of a vestigial organ ; 

 hence we conclude that these bodies are vestigial, consequently have 

 been more important than they now are, and that the origin 

 primitively was from marginal tentacles. As bearing upon this, it is 

 important to remember that it is admitted generally that tentacu- 

 locysts arise from modification of ordinary tentacles. 



These facts bring us now to the conclusion that Halidystus is 

 descended from an animal possessing a stage wherein certain marginal 

 tentacles had become modified into highly important sensory organs, ap- 

 proaching, if indeed not almost similar to, the tentaculocystsof medusae. 

 Further, organs of this complex description can scarcely be imagined 

 as in any way serviceable to a sedentary animal unprovided with 

 powers of locomotion, whatever use such may subserve in a free- 

 swimming organism, whether we follow the old theories adducing 

 auditory and visual functions to these sense-organs, or follow Dr. 

 Hurst's more probable theory as put forward recently (3). Sessile 

 animals of the description of the Lucernariae can neither follow prey 

 nor avoid capture, and hence do not require auditory and visual 

 organs, much less automatically-actuated direction-controlling bodies. 

 This brings me back to my initial proposition, that among the 

 ancestors of Halidystus was a form having a strobilation stage throwing 

 off ephyrae, which became acraspedote medusae provided with eight 

 marginal sensory organs and with numerous tentacles separating 

 these, approaching indeed such a form as the present-day Amelia. 

 Then, when the premature development of sexual glands occurred in 

 the scyphistoma stage, as already mentioned, it is reasonable to infer 

 that the other organs belonging normally to the fully-developed adult 

 form (acraspedote medusa) would also appear. Chief among these 

 would be the marginal sensory bodies ; but these would occupy a 

 very different platform to the sexual glands. The latter are necessary 

 to the existence of the animal ; the former — under the changed 

 conditions of sexual maturity in sessile position — would be of no more use, 

 as pointed out above. The courses open to them must either be, (a) 

 reversion to the original form, i.e., that of marginal tentacles, (&) 



