482 



SCIENCE. 



I N. S. Vol. VIII. No. 197. 



resemblance to the Hydi'omedusse is regarded 

 by most naturalists as one of the numerous 

 cases of convergent evolution exhibited by the 

 two groups of jelly-fish (Hydro- and Scyphome- 

 dus8e), due to similarity in environment and to 

 a certain similarity in the ancestral polyps fi'om 

 which the two groups have been derived. 



The Cubomedusse are so rare that in spite of 

 their interesting features, interesting alike to 

 the student of phylogeny and nerve-physiology, 

 few naturalists have had the opportunity of 

 studying them. Our knowledge of the group 

 has rested mainly on Claus's description of 

 Charybdea marsupialis (Wien. Arb. 1878). This 

 very valuable paper, as Conant remarks, is 

 written in a style difficult of comprehension, 

 and many students who read with pleasure and 

 profit the lucid treatises on medusan structure 

 by the Hertwigs and Haeckel have turned 

 away discouraged from Claus's work. To 

 Claus's account, Haeckel in his ' System' has 

 added but little. The only other investigator 

 the group of is SchewiakofF (1889), who has 

 studied the remarkable sense organs. 



Through Conant' s discovery in 1896 of two 

 new species (Charybdea Xaymacana and Tripe- 

 dalia cystophora), which are present in abun- 

 dance in Jamaican waters, the Marine Labora- 

 tory of the Johns Hopkins University has once 

 again made accessible to students, material for 

 the pursuit of investigations of wide interest. 

 It was for the purpose of continuing his study 

 of this group that Conant, in the summer of 

 1897, revisited Jamaica, and, as we learn from 

 Professor Brooks's introduction, he succeeded 

 in making many observations on the physiology 

 of the sense-organs and on the embryology. 

 His notes and material, we are told, are in 

 such shape that they can be handed over to 

 some one else, and it may be safely predicted 

 that a valuable contribution to science will be 

 the outcome of the last summer's work of this 

 talented young naturalist. 



The account of the subomedusa structure 

 given by Dr. Conant is succinct, but compre- 

 hensive. The deep, four-sided bell bears a ten- 

 tacle (or in some species a bunch of tenta- 

 cles) at each angle. On each lateral surface, 

 at a higher level than the tentacles, is situated 

 a niche into which projects a sense-organ. The 



primitively undivided (Scyphistoma condition) 

 gastrovascular space is here differentiated into- 

 a central stomach and a peripheral portion 

 lying in the lateral wall of the bell. The 

 peripheral portion is subdivided into four stom- 

 ach pockets by linear partitions, lying in the 

 plane of the tentacles and therefore interradiaL 

 These partitions (cathammse) are mere strips of 

 entodermal lamella, produced by the fusion be- 

 tween the entodermal linings of ex- and sub- 

 umbrella. The cathammal lines stop short of 

 the tentacles, leaving an undivided peripheral 

 portion of the primitive space, by means of 

 which the four stomach pockets communicate 

 with one another. As Conant points out, the 

 arrangement recalls the gastrovascular system 

 of many HydromedusiE, with the difference that 

 in the Cubomedusse the radial canals are wide 

 ' stomach pockets' and the cathammal plates- 

 are narrow lines. When we come, however, to 

 the extreme peripheral portion of the gastro- 

 vascular system, we find that the likeness is not 

 with the Hydromedusae, but with the lobed 

 ScyphomedussB. The gastrovascular space, to 

 be brief, does not end with an even circular 

 edge at the bell margin, as is the rule in the 

 former group, but is divided into separate lobes 

 (marginal pockets) extending into the velum 

 (as velar canals). Conant does not dwell on. 

 phylogenetic inferences, but evidently inclines 

 to the belief that the ancestors of the Cubome- 

 dusse possessed a margin divided into sixteen 

 lobes. The present position of the four sense 

 organs indicates the site of the original margin, 

 "which elsewhere has grown down and away 

 from its former level, leaving the sensory clubs 

 like floatage stranded at high-water mark."' 

 Fusion between adjacent lobes, involving the 

 ectoderm and jelly, gave to the medusa a con- 

 tinuous margin and a ' velum,' but, owing to the- 

 incompleteness in the fusion of the entodermal 

 linings of the several lobes, the latter still re- 

 tain in the adult Cubomedusa enough of their 

 individuality to indicate their former condition. 

 In a word, the marginal pockets of the existing 

 Cubomedusa are to be construed as entodermal 

 linings of once separate lobes. 



This conclusion as to the morphology of the 

 marginal pockets derives much support from 

 the behavior of a puzzling structure, called by 



