Alternation of Generations in the Medusae. 441 



nina only on those days when the Geryonia also appeared in the 

 Gulf of Nice, but then always accompanying the latter in great 

 quantities. 



The youngest individuals of Cunina rhododactyla that I have 

 observed, of which the umbrella was 3 millims. in diameter, 

 agreed in all essential particulars so closely with the oldest 

 observed Geryonia-h\xdi^ of 1 millim. diameter, that I can no 

 longer doubt the identity of the two forms. As in the bud of 

 the Geryonia, the thick disciform umbrella is divided at the 

 margin by eight deep notches into the same number of lobes, 

 each of which bears at its apex a pedunculate free sensory vesicle. 

 Eight similar tentacles are attached in the notches. The simple 

 unpedunculated stomach gives off from its circumference eight 

 flat and broad radial sacs, which reach to the base of the tenta- 

 cles, and are there united by a narrow annular vessel which runs 

 along the border of the marginal lobes. The only difference, 

 besides the smaller size and plumper form of the stomachal buds 

 of Geryonia, that I can find between the oldest of these buds 

 and the youngest individuals of Cunina consists in the fact that 

 the tentacles of the latter are more slender and longer, and, on 

 the other hand, the stomach is flatter and shorter — differences 

 which would, no doubt, disappear by the observation of the in- 

 termediate age 2 millims. in diameter. 



I have also been able to ascertain the further development of 

 Cunina rhododactyla to full sexual maturity. It consists essen- 

 tially in the gradual increase of the number of similar segments 

 composing the body from eight to sixteen, a new segment being 

 inserted from time to time between those previously existing. 

 The oldest animals have an umbrella 10-11 millims. in diame- 

 ter, and possess sixteen tentacles, sixteen stomachal sacs, sixteen 

 marginal lobes, and a great but indeterminate number (between 

 50 and 100) of sensory vesicles. The latter increase in a very 

 irregular manner, so that the different marginal lobes of one and 

 the same animal bear from four to eight vesicles. The sexual 

 products are developed in the lower wall of the stomachal sacs, 

 from its epithelium. 



From all that has been stated, it appears to me to be no longer 

 doubtful that the octoradiate buds which sprout from the tongue 

 of the sexually mature sexradiate Geryonia [Carmarind) hastata, 

 within its stomachal cavity, are really developed directly into the 

 sexually mature animal of Cunina rhododactyla. Should this 

 supposition (which I cannot but regard as a certainty) be con- 

 firmed, it requires nothing further to show that we have here an 

 exceedingly wonderful and perfectly and fundamentally new 

 form of the alternation of generations, if indeed we may apply 

 that name to this singular process. It might better be called 



