Budding in Perophora. 219 
which agree neither with those of Hjort on Botryllus* and 
Glossophorum f, of Hjort and Bonnevie on Distaplia f, nor 
with the results obtained*by Ritter on Perophora §. 
In all of the above-mentioned Ascidians except Perophora 
the rudiment arises as an evagination of the dorsal wall of the 
inner vesicle, and in Perophora, according to Ritter, by 
migration of cells from the dorsal portion of the endodermal 
vesicle into the rudiment. For the reasons given below I 
believe that in Perophora viridis the structures in question 
have a common mesenchymatous origin. 
The first appearance of the rudiment is at a stage when the 
rotation of the endodermal vesicle is about half accomplished 
and before there is any indication of the fold which is to give 
rise to the right peribranchial sac. 
In the anterior region of the bud a little to the left of the 
median dorsal line there is seen a slender elongated mass of 
cells loosely grouped together and lying against the outer. 
surface of the inner vesicle. ~* 
Although I have examined my sections with the greatest 
care, I have failed to find any indication of cell-migration 
from the endodermal vesicle, and hence cannot confirm Ritter’s 
statements that an “ indistinguishable transition from the cells 
of the ‘ endoderm’ to those of the neuro-hypophyseal anlage 
is to be traced,” and also that cells can be traced ‘in the act 
of migrating from the endoderm into the anlage.” 
In the first place the line of separation between the rudi- 
ment and the vesicle is seen with perfect clearness to be 
everywhere intact, and, furthermore, the cells of the rudiment 
are identical in appearance with the free mesenchyme-cells ; 
the latter are, moreover, more numerous in this region than 
elsewhere. During the early stages of development, that is, 
before the rudiment has become very compact, the similarity 
between these cells is very evident (fig, 4, A). I therefore 
conclude that, just as in the case of the pericardium, the 
common rudiment of the dorsal tube and nerve-ganglion is 
formed by free mesenchyme-cells, 
The cells of the rudiment are at first loosely and irregularly 
piled on one another ; but soon the mass becomes more closely 
packed and assumes a cylindrical form (fig. 4, B). Other 
mesenchyme-cells are added to the string from the outside 
and active cell-division goes on within the mass. 
When the rotation of the inner vesicle is completed the 
nerve-rudiment has been carried up from the left side to the 
* Mitth. Zool. Stat. Neapel, x. Bd. 
+ ‘Anatomischer Anzeiger,’ x. Bd. no. 7. 
t Ibid. no. 12. § Ibid. no. 11. 
