Heredity and Natural Selection. 305 



two larval forms have diverged in two directions from 

 this ancestor, from which they inherit all that they have 

 in common. 



ILieckel believes that after this separation took place, 

 the veiled medusae were developed from hydroid polyp?,, 

 while the veilless medusae were developed from scyphos- 

 toma polyps. The many points of resemblance between 

 the two forms of medusae are, therefore, not due to 

 common inheritance, but have been secondarily ac- 

 quired. They are due to the fact that the two groups of 

 medusae have been evolved along parallel but distinct 

 lines. 



HaeckePs familiarity with the medusae entitles him to 

 speak with great authority; but still he may possibly be 

 wrong, and the origin of the two groups may not have 

 been as he supposes. 



There are four possible hypotheses as to the origin 

 of the medusae, in favor of each of which something may 

 be said. We may hold with Haeckel that the two larval 

 polyps are the divergent descendants of a common an- 

 cestral polyp, which had no medusa stage, and that 

 each has subsequently developed medusae, or we may be- 

 lieve that the common ancestor was a medusa without 

 a polyp larva, and that the hydra larva and the scy- 

 phostoma larva have been independently acquired, or 

 we may believe that the ancestral form had both a lar- 

 val polyp-like stage, and an adult medusa stage, or fin- 

 ally we may assume what seems to us the most probable 

 view, that the ancestral form was neither a true swim- 

 ming medusa nor a true sedentary polyp, but some- 

 thing half-way between, like the actinula of Tubularia 

 or the embryo of Turritopsis. I do not see any fifth al- 

 ternative, and one of these four suppositions must cor- 

 respond with the actual evolution of the group. Now 



