.\\THuMi:i>[ S.K HnrCIATXVII.i.IA. 1C.7 



derived at an earlier stage from the entoderm of the parent medusa and have migrated into 

 the ectoderm. Careful search has, however, failed to reveal evidence of any such migration, 

 and in default of evidence to the contrary, we assume that these cells niiiv he ectoilermal in 

 origin, although destined to produce the entoderm of the hud. 



The central cavity of the vesicle is partially rilled with a loose mass of nuclei and cell ma- 

 terial showing little or no trace of cell boundaries (r, plate 18, fig. 2). These are present at all 

 stages of the developing hud, partially filling the gastrovascular cavity, hut they appear to 

 degenerate in later stages, and it seems possible they serve to nourish the bud in its urowth. 

 This supposition appears the more probable from the fact that the gastrovascular ca\itv of the 

 bud is never in connection with that of the parent medusa. 



\Ve have, therefore, a graded series of phenomena in the asexual production of medusa- 

 buds by hyd mined usx. Where the ectoderm and entoderm are both thin-layered and of about 

 equal thickness, such as in Stirs in and F.uclifilr,t,i p,nnj<,\iia, both entoderm and ectoderm 

 take an equal share in the formation of the bud. In forms such as Ruthk,,: 'i.t't[>un<-tntti and 



* ' ' 



1 i I 



1 . i m 



S: 



-V Ao,," 



Sq. 89. 9- 



l-'n,. Sij. Hvilrohl of Hvi<i!<i:niillni carolinfmis, after A. Agassi/, in North American Acalepluf, 



showing an eiilar^eil view of .1 part of >tem liearin^ medusa-buds. 

 I']'.. 90. Boutylim'illiil niuhf, from life, liy the author. Tortuy.is. I ; lon.la, June 2", 1908. 



Lizzin fliiftrirt-tli-i, according to Chun, Bibliotheca Zoologica, 1895, a different condition is 

 observed, for the medusa-buds are formed entirely within the ectoderm of the parent, although 

 the gastiovascular cavity of the bud finally acquires a connection with that of the parent, tin 

 entoderm of the bud becoming continuous with that of the parent manubrium. 



In Bougainvillia nin/ii-. however, the ectoderm is very thick and the budding medusae 

 are developed within it alone, the entoderm of the parent medusa remaining inert and passive 

 during the growth of the bud, and no connection ever being established between the gastro- 

 vascular cavities of the bud and the parent. This result may he regarded as due to a gradual 

 process favored by the thickness of the ectoderm, which prevented the deep-lying entoderm 

 from taking an equal share in the formation of the bud. until fiuallv it remains passixi- 

 throughout the period of the formation of the bud. Medusa? produced from ectoderm alone 

 may, therefore, be phylogenetically homologous with medusae produced by the more primi- 

 tive and universal cooperation of both ectoderm and entoderm. 



The entoderm of the manubrium and tentacle-bulbs of K. niobe is rosin-colored, all other 

 parts being transparent. There is a dark-colored ocellus on the velar side of each tentacle 

 near the basal bulb. This medusa is found in Nassau Harbor, Bahamas, in April. 



