578 



MEDUS.E OF THE WORLD. 



Hadzi finds that some of the free-swimming planulae of Chrysaora are 4 or 5 times as large 

 as others. They swim with the broad end forward and soon settle upon nlva, etc., attaching 

 by means of their forward ends. The entoderm, which was previously a solid mass, then 

 hollows out and the larva becomes two-layered, and the uppermost (the former posterior) end 

 becomes the widest. The mouth then breaks through, the oral pole flattens laterally, and 4 

 tentacles develop, 2 in the short and 2 at the ends of the long diameter. The stomach-pouches 

 do not begin to form until after the mouth and 4 tentacles have developed. 



The view of Claus has received strong support from Hadzi who casts serious doubt upon 

 Geotte's interpretation that the oesophagus of the scyphostoma is always composed of invag- 

 inated ectoderm. 



Hadzi, whose research upon Chrysaora appears to have been carefully studied, finds that 

 the oesophagus of the scyphostoma is entodermal and that the mouth breaks through from the 

 inside, the entoderm thus taking the active share in its formation, and no invagination of ecto- 

 derm occurring. Indeed Hadzi finds that the cells lining the throat of the scyphostoma resem- 

 ble ectodermal cells in having nematocyst capsules and glands, but they are nevertheless solely 

 of entodermal origin. From this it follows that the 4 primary stomach-pouches are also ento- 

 dermal, not 2 ectodermal and 2 entodermal as claimed by Goette. Hadzi finds also that the 4 

 intertentacular taeniolae are formed from 4 simple, longitudinal infoldings of the entoderm of 

 the stomach wall, the ectoderm taking no part in their formation. The primary stomach- 

 pouches are thus the passive result of the infoldings which form the taeniolae, not of an active 

 outgrowth of pouches as Goette believes. 



Hadzi's view appears to be the more reasonable, for if Goette were correct one half of the 

 gonads would be ectodermal and one half entodermal, whereas according to Hadzi they are all 

 entodermal; moreover, according to Goette, the mouth of the first ephyra set tree in strobili- 

 zation has its oesophagus lined with ectoderm, while those ephyrae which follow it have their 

 throats lined with entoderm, an anomalous condition. According to Hadzi and Heric, how- 

 ever, all of the ephyrae have their throats lined with entoderm. 



Heric finds in the stabilization of the scyphostoma of 

 Chrysaora that with the exception of the terminal ephyra all of 

 the mouth-tubes of the chain of ephyrae are formed from the 

 connecting tube which joins all of the ephyrae together. The 

 external wall of this connecting tube is ectodermal and its inner 

 wall entodermal. 4 perradial clefts develop in the side wall of 

 each tube near the upper end where it joins with the exumbrella of 

 the overlying ephyra. The lower edges of these clefts grow out- 

 ward and form the 4 lips of the ephyra, while the 4 connections 

 are interradial and are in the radii of the taemolae which consti- 

 tute their inner sides. 



The 4 subgemtal cavities of the ephyra are new formations 

 and not derived from the 4 funnel-cavities of the scyphostoma. 

 The 4 interradial septa of the stomach-cavity of the ephyra are, 

 however, derived from the taemolae of the scyphostoma. These 

 soon disappear, and the central stomach of the medusa is a 

 simple lenticular space. 



The forms of Chrysaora are so imperfectly separated one from 

 another that were it not for the fact that many minute distinc- 

 tions have been pointed out between them, I would greatly 

 prefer to consider them all to be one variable species, C. h ysoscella. 

 However, we may possibly distinguish more or less vaguely: 



FIG. 365. Diagrammatic section of 

 a strobilla of Chrvsaora after 

 Heric, in Arbeit. Zool. Inst. 

 Wien. 



ah, perradial cleft, am, beginning of 

 the lip. />, /, /, stages in the 

 growth of the throat-tube. 

 s m, septal muscle. Ectoderm 

 cross-hatched, entoderm plain, 

 intermediate lamella dotted. 



C. hysoscelIa=C mediterranea with its varieties biossevillei and plocamia (?) 



of the Atlantic, Mediterranean, and South Pacific. 

 C. heh'ola, with its varieties calliparea, and cliinensis of the Pacific and 



Indian Oceans. 

 C. melanaster with its variety gilberti of the North Pacific. 



I believe that a study of the following synoptic table will 

 convince one that we have here only one species, the varieties of 

 which defy classification in terms of the Linnean system. 



