THE BUDDING AND FISSION OF ASCIDIANS. 525 



(u') has already taken on the character of an ovum, and is 

 surrounded by a rudnnental ovisac. The conical bud elongates 

 and dilates at its extremity, and the dilatation gradually takes 

 on the form of a new zooid united by a narrow neck, or pe- 

 duncle, with the parent (Fig. 150, IV.). The endostylic cone 

 gives rise to the whole alimentary canal of the bud, while the 

 ectoderm of the latter proceeds from the ectoderm, and its 

 ovisac and testis from the mesoblastic cells, of the parent. 

 Thus the organs of the bud are all the direct product of the 

 corresponding parts, or of the primitive layers of the germ 

 from which they are derived, in the parent.^ 



After the terminal bud is formed, a second is usually de- 

 veloped immediately below it (Fig. 150, VI.) by the growth 

 of the ectoderm, endodermal axis, and mesoblastic cells of the 

 peduncle ; and it would appear that this process is frequently 

 repeated. The fully-formed bud becomes detached, and takes 

 its place among the other zooids in the test, there to repeat 

 the process of gemmation. 



The observations of Krohn, MetschnikofF, and Kowalewsky, 

 have shown that two components enter into the buds of ascid- 

 ians in general ; first, an outer layer consisting of the ecto- 

 derm of the region in which the budding takes place, and, 

 secondlv, an inner layer derived from the endoderm of the 

 branchial sac (Pe7^op/iora) ; or, as in J3otn/llus, according to 

 Metschnikoif, from the atrial tunic. ^ To these must be added 



^ In my second memoir on Pyrosoirna ('' Trans. Linn. Society," xxiii., p. 

 211) I have said : 



" Gemmation does not take place in Pi/rosoma as in so manv of the lower 

 animals (e. g., the Bydrnzoa and Polyzoa^ or Saipa and Clavelina^ amonsr the 

 ascidians), by the outgrowth of a i»rocess of tlae body-wall, whose primarily whol- 

 ly indifferent parietes become clitferentiated into the organs of the bud; but, 

 from the first, seyeral components, derived from as many distinct parts of the 

 parental organism, are distins^uishable in it, and each component is the source 

 of certain parts of the new being, and of them only. Thu"^ the bodv-wall or 

 external tunic of the parent gives rise to the external tunic of the bud ; while a 

 process of the endostvlic cone of the parent is converted into the alimentary 

 tract of the bud, and the reproductive ororans of the Litter are furnished by a 

 part of that tissue v»^hence the reproductive organs of the parent took their 

 origin." 



As will appear further on. however, recent investigations show that the 

 ■whole process of budding in the frreat majority of the Tiivicata^ and at any rate 

 the first steps of that process in Salpn, are essentially similar to those in Pyr^o- 

 snma ; and it remains to be seen whether there is any difference in other As- 

 cidians. And as reofards even the Tlydrnzoa, the expression that the parietes 

 of a bud are at first " wholly indifferent " in structure is not quite accurate, in- 

 asmuch as they are composed of an ectodermal and an endodermal layer, which 

 are continuous with those of the parent, and give rise to homologous organs. 



2 If, as some observations tend to show, the atrial tunic itself is a diver- 

 ticulum of the primitive endoderm, this case would form no exception to the 

 general law of budding in the Tunicata. 



