514 TUNICATA. 



the barrel-shaped " nurse," the lateral and the median buds of differ- 

 ently formed individuals of one and the same stock (pp. 388 and 474). 



In Pyrosoma, the alternation of generations consists of successive 

 generations of Cyathozooids and Ascidiozooids. The Cyathozooid 

 corresponds to the solitary form of tialpa, but remains in an un- 

 developed condition, its development being restricted to the embryonic 

 period. It reproduces itself only asexually. The Ascidiozooids, on 

 the contrary, besides multiplying sexually, have retained the capacity 

 for budding, and therefore, in the cycle of generations of Pyrosoma, 

 one generation (Cyathozooid) produced from the egg, alternates with 

 several generations (Ascidiozooids) which have arisen through bud- 

 ding. 



The alternation of generations which takes place in the Doliolidae 

 has already been described in detail (p. 474). 



The alternation of generations in the Tunicates must therefore be 

 derived from colony-formation and must be regarded as a case of 

 division of labour (LEUCKART). The only difficulty that remains to 

 be explained is in what way the hypothetical ancestors of the 

 Tunicates which became attached acquired the capacity for repro- 

 duction by budding. It is not easy to understand how an animal 

 which had always reproduced itself sexually should come to produce 

 buds. There is considerable evidence tending to show that, in the 

 Tunicates, budding developed out of fission. The asexual repro- 

 duction of Amaroimmn especially must be regarded as an actual 

 process of division (p. 449). We are therefore justified in assuming 

 that the hypothetical ancestors of the Tunicates, besides their sexual 

 multiplication, at first reproduced themselves through fission, and 

 that the later budding and stolon-formation developed out of this 

 manner of reproduction. 



BALFOUR * and ULJANIN (No. 86) have tried to remove the 

 difficulty which arises if we regard the capacity for multiplication 

 by fission as acquired only after attachment, by suggesting that this 

 process appears first in the embryos. According to this view, the 

 Tunicates first acquired a capacity for dividing in the first embryonic 

 stages, as is the case in Lumhricus trapezoidvs (Vol. I., p. 281), a 

 capacity which was only secondarily passed on from the embryonic 

 stages to the adult form. But when we remember the great capacity 

 for regeneration possessed by the Tunicates (p. 448) we shall hardly 

 find it necessary to fall back on such an hypothesis. There is no 



* Text-book of Gomp. Embr., Vol. II., p. 34 (footnote) 2nd edit., 1885. 



