LIFE HISTORY STUDIES OF ANIMALS. 497 



natiou of budding with egg production and the unlikeness xU form of 

 the asexual and sexual stages. Like Ohamisso, he carefully distin- 

 guished between development with metamorphosis and alternation of 

 generations. All three naturalists, Chamisso, Sars, and Steenstrup, 

 laid stress on this point. In an insect, they would have said, there is 

 development with metamorphosis. The same animal passes from larva 

 to i)upa, and from pupa to imago. In Aurelia or Salpa, however, the 

 animal which lays eggs is not the animal which buds, but its progeny. 

 Tlie cycle of the life history includes two generations and many 

 individuals. 



This view has spread very Avidely, and if we were to judge by what 

 is commonly taught I think that we should recognize this as the doc- 

 trine Tiow ])revalent. It is, however, in my opinion, far inferior as an 

 explanation of the facts to that adopted by Leuckart, Carpenter, and 

 Huxley, Avbo regard the whole cycle, from egg to egg, as one life 

 history. Huxley and Carpenter, diifering in this from Leuckart, do 

 not shrink from calling the whole product of the egg an animal, even 

 though it consists of a multitude of creatures which move about and 

 seek their food in complete independence of one another. Eather than 

 ignore the unity of the life history of Aurelio or Salpa, they would 

 adopt the most paradoxical language. This attitude was forced upon 

 them by the comparative method. They refused to study Aurelia, for 

 example, as an animal apart. It had its near and its remoter relatives. 

 Among these is the fresh-water Hydra, which develops without trans- 

 formation, buds off other hydras when food is plentiful, and at length 

 becomes sexually mature. Budding is here a mere episode, which 

 may be brought in or left out, according to circumstances. The same 

 individual polyp which buds afterwards produces eggs. The life 

 history of Salpa can not be traced with equal facility to a simple 

 beginning, for it presents points of difficulty on which the learned 

 differ. In the polychait worms, however, we find a beautiful gradation 

 leading up to alternation of generations. We begin with gradual 

 addition of new segments and increasing specialization of the two ends 

 of the body, the fore end becoming nonreproductive and the hinder 

 end reproductive. Then we reach a stage (syllis) in which the repro- 

 ductive half breaks off from the fore part and forms (after separation) 

 a new head, while the fore part adds new segments behind. In Auto- 

 lytus the new head forms before separation, and many worms may 

 cohere for a time, forming a long chain with heads at intervals. In 

 Myrianida the worms break up first and afterwards become sexually 

 mature. We should gather from these cases that alternation of gen- 

 erations may arise by the introduction of a budding stage into a devel- 

 opment with transformation. The polyp or worm buds while young 

 and lays eggs at a later time. The separation of the two processes of 

 reproduction often becomes complete, each being restricted to its own 

 place in the life history. As a rule the worm or polyp will bud while 

 SM 97 32 



