.500 MEDUSA OF THE WORLD. 



been acquired independently of and is not homologous with the bell of the Leptolina medusae. 

 (See Goette, 1907, Zeit. fiir wissen. Zool., Bd. 87, p. 289.) 



The peculiar velum of the Carvbdeidae among the Scyphomedusae is a structure of the 

 subumbrella, not of both subumbrella and exumbrella as in H\dromedusae. It may be regarded 

 as a parallelism, and not genetically related to the velum of Hydromedusae. The exumbrella 

 nerve-ring, found commonly in Hydromedusae, does not exist in the Scyphomedusae, but is 

 replaced bv a subumbrella plexus of fibers extending between the marginal sense-organs and 

 also radialh" inward from these ganglionic centers. There is also a diffuse, nervous, epithelial, 

 ectodermal network over the subumbrella. The subumbrella alone is sensitive to stimuli, the 

 exumbrella being non-sensory, but covered with a nematocyst-bearing epithelium. 



In the Scyphomedusae the mature, sexual products are found in the entoderm, whereas 

 the\' are usualh' in the ectoderm in H\dromedusae. The velum which is universalh' present 

 in H^dromedus.x is absent in Sc\phomedusae. The 4 interiadial, gastric septa which are 

 always seen in the scyphostoma larva of Scyphomedusae are not found in the polyp stage of 

 Hydromedusae. 



\\ ben we come to consider the relationships of the several orders constituting the Scypho- 

 medusae themselves we meet with difficulties which render our classification onlv tentative. 

 The history of the attempts to classify the Scjphomedusae have recently been reviewed by Maas, 

 1907 (Ergeb. und Fortschritte der Zool., Bd. i, p. 189), and by Bigelow, 1909 (Mem. Museum 

 Comp. Zool. at Harvard College, vol. 37, p. 13). Marked advances over Haeckel's artificial 

 classification were made bv Claus, 1878, 1883, who showed that the Car\ bdeida are a very 

 aberrant group not closely related to other Sc\'phomedusae, and not descended from the Stauro- 

 medusae as Haeckel believed. As Claus showed, the interradial, gastric septa in Car\bdeidae 

 may be newly arisen fusions of the entodermal walls of the exumbrella and subumbrella and 

 not derived from the interradial tsniolae of the sc\phostoina. Untortunateh', however, no one 

 has seen the late sc\phostoma stage in Car\bdeidac, and hence our knowledge of their rela- 

 tionships must remain doubtful. Claus, 1883, and Vanhoffen, 1892, called attention to the 

 supposed importance of the presence or absence of the interradial, gastric septa in the classi- 

 fication of sexually mature Sc\phomedusae but we must remember that these are absent in 

 Semffostomac and found in Coronatae, yet these two orders are closeh' related. Moreover, 

 interradial gastric septa are found both in Stauromedusa? and Car\bdeida; yet all modern 

 students agree that these groups are widely divergent. The sessile Stauromedusae certainly 

 approach very closely in structure to the tvpe of the scyphostoma, but this may be the result 

 of degeneracy due to their sedentary habit. 



Vanhoffen, 1892, in a masterly paper called attention to the importance of the annular 

 furrow and marginal, exumbrella sculpturing of the bell in certain of Haeckel's Discome- 

 dusae, and he grouped these under the designation Coronatae. Claus, 1883, had previously 

 shown that in N aiisitho',', which is a t\pical member of the Coronatae, the interradial septa of 

 the central stomach are derived from the 4 taeniolae of the sc\phostoma. These septa are, how- 

 ever, absent in the medusa-stage in Semaeostomeae and Rhizostomac, although they are found 

 in the scyphostoma. 



All evidence both anatomical and cmbrvological shows that the Rhizostomae are descended 

 directly from the more simply organized Semaeostomeae. It is probable that the Coronatae and 

 Semaeostomeae are not descended one from the other, but are derived from some common 

 ancestral stock which has disappeared. The Stauromedus.'e may represent a highly specialized 

 derivative from this common ancestral form, their specialization being due to their sessile, 

 attached habit of life. The Carybdeidx are so aberrant and our knowledge of their develop- 

 ment is so imperfect that we have not yet been able to determine their relationships to other 

 Scyphomedusje. 



The habits and distribution of the several orders of ScA'phomedusae are characteristic. 

 The Stauromedusae are found only in the Arctic and Antarctic regions and in cold seas, 

 and are unknown from the tropics; thus furnishing us with an interesting problem in dis- 

 tribution. The planula is not pelagic but creeping and is devoid of cilia, and the medusae 

 are sessile or creeping, and do not pulsate rh\thmicalli,'. They affix themselves to seaweeds 

 and rocks and rarely move from their places of attachment. They are probably degenerate 

 forms although their relationship to other Sc\phomcdusac is uncertain. They are confined to 

 the shores, but may be carried far from their normal habitats upon drifting seaweed. 



