Alternation of Generations in the Medusa. 439 



globular larva first of all developes six solid radial subsidiary- 

 tentacles, then six solid interradial tentacles, and afterwards six 

 interradial sensory vesicles. Then only do the six hollow radial 

 primary tentacles make their appearance, and lastly, after these, 

 the six radial sensory vesicles. The first twelve solid tentacles 

 are then lost, and only the last six, the hollow radial primary 

 tentacles, remain. At the same time the long stomachal pedun- 

 cle, which was at first entirely wanting, is developed. 



These sextuple larvse, the metamorphosis of which into the 

 fully-developed Geryonia hastata may be traced through all its 

 stages, are probably the products of sexual reproduction. But 

 the same animal also developes young Medusas asexually, and, 

 indeed, by gemmation in the interior of the digestive stomachal 

 cavity ; and these have a perfectly different form and structure. 

 These Medusoid-buds are probably the same that Krohn saw 

 on one occasion. But they are not, as stated by him, sextuple, 

 nor do they become developed into a Geryonia ; but they are 

 octuple, and are developed into a totally different Medusa, very 

 probably into a species of the family JEginida, described by me 

 as Cunina rhododactyla. 



This gemmation, which is exceedingly remarkable, both on 

 account of its locality and of its heterogeneous product, occurs 

 only in the stomachs of sexually mature animals, and in 

 both sexes. I was able to examine twenty-three individuals of 

 Geryonia hastata with regard to the conditions of this pheno- 

 menon. Of these no less than nine had the stomach reduced 

 to a stump, or in course of reproduction. Of the other fourteen, 

 seven showed a long spike of closely united eight-rayed buds in 

 the stomach ; of these seven animals three were males, and four 

 females, all with perfectly mature sexual products in the genital 

 leaves. The number of buds in the stomach of each animal 

 varied from twenty to nearly a hundred. The buds were seated 

 close together, with the vertical surface of the umbrella (the 

 aboral pole) attached to a long cylindrical process which was 

 fixed to the base of the stomach. This process is nothing but 

 the long dagger-like prolongation of the stomachal peduncle, 

 which, in Carmarina, as in Glossocodon, projects freely into the 

 stomachal cavity, and in non-gemmiparous animals is often 

 protruded from the mouth, and seems to subserve the function 

 of a tongue. In two of the largest Geryonia I counted the buds 

 attached to the tongue and forming with it a thick cylindrical 

 spike, which hung down freely in the middle of the campanulate 

 stomach like the clapper of a bell. One spike was composed of 

 seventy-one, the other of eighty-five buds. Young and old buds, 

 in the most various stages of development, are seated indiscri- 

 minately together. 



30* 



