KEPOET 0 ^ THE TUNICATA. 
389 
the accompanpng diagram (fig. 11) by the line stretching from the splitting point of 
the Proto-Timicata to the most recent common ancestor of Simple and Compound 
Ascidians (A. in diagram). 
This latter form was probably something like a Clavelina without a peduncle. • It 
was attached by the posterior end of the body, and was elongated antero-posteriorly. 
The branchial aperture was at the upper end, and the atrial far forward on the dorsal 
edge, and both were circular and T\dthout lobes. The branchial sac was not very 
large, and consequently the greater part of the alimentar)" canal lay posterior to it. No 
folds were present in the branchial sac, and the stigmata were straight. The tentacles 
were simple. Finally, by reproducing by gemmation, small colonies were produced. 
The buds were probably formed as simple outgrowths (containing some cells derived 
from the endoderm of the parent covered by a prolongation of the body- wall) from the 
posterior end of the body. 
From this form (A. in diagram, fig. 11) several lines diverged. One led to the genus 
Clavelina, in which little change has taken place. The body has become elongated, and 
more or less pedunculated, while the buds have come to be produced on stolons which 
are prolongations of the body- wall of the parent Ascidiozooid covered by a layer of test 
and containing blood sinuses. A second line of descent, starting from A., has produced 
the genus Perophora, in which the body has become shortened antero-posteriorly, so that 
the ahmentary canal has come to lie alongside the branchial sac. The third diverging line 
from A. is that representing the ancestors of the remaining Simple and Compound Ascidians. 
In these forms the branchial sac became complicated by the addition of internal longitudinal 
bars which were not previously present. The genus Ecteinascidia ^ is closely related to 
these ancestral forms. 
At about this point (B. in fig. 11) the line split into two great series, the one leading 
to the more typical Compound Ascicbans (the Polyclinidae, the Distomidae, &c.) and to 
Pyrosoma, and the other giving rise to the typical Simple Ascidians, and to the 
Botryllidse and the Poly sty elidse. The first of these two important branches soon divided 
(at C. in fig. 11) into two lines of descent ; the one leading to the Polyclinidse, and the 
other through the Distomidae to the Didemnidae, the Diplosomidae, the Coelocormidae, and 
finally the Pyrosomidae. In both of these lines, and in their common ancestors from the 
point B. onwards, the power of reproducing by gemmation was retained and even 
increased, and the members of the resulting colonies became more closely ’^united with one 
another than is the case in Clavelina and allied forms. 
The line which leads from C. to the existing Polyclinidae must have been occupied by 
a series of forms in which the body of the Ascidiozooid became more and more elongated 
antero-posteriorly, and finally divided more or less distinctly into three regions, the 
1 See Part I. of this Eeport (vol. vi., 1882), p. 239 ; and Sluiter, Ueber einige einfachen Ascidien, &c., Natuurkund. 
Tijdsch. V. Nederland. Indie, Bd. xlv. p. 160, 1885. 
