30 
THE VOYAGE OF H.M.S. CHALLENGER. 
Milne-Edwards classified the genera of the Compound Ascidians in three “ tribes,” the 
“ Polycliniens,” the “ Didemniens,” and the “ Botrylliens,” an arrangement which with 
our present knowledge of the group still seems fairly natural in some respects. These 
three divisions are distinguished by such anatomical characters as the relations of the other 
viscera to the branchial sac. In the “ Polycliniens ” the body has three regions — the 
“ thorax,” containing mainly the branchial sac ; the “ abdomen,” formed chiefly of the 
stomach and the greater part of the intestine ; and the “ post-abdomen,” having the 
reproductive organs £ind the heart. In the “ Didemniens ” there are only two regions — 
thorax and abdomen — the reproductive organs and heart being placed on the intestine. 
In the third group, the “ Botrylliens,” the viscera form a single mass in which the 
alimentary canal lies alongside the branchial sac. The flrst and last tribes are natural 
groups, and correspond respectively to the modern families Polyclinidse and Botryl- 
lidse, but the second tribe, the Didemniens, is rather too wide in its range, and includes 
forms which are not closely related. Milne-Edwards, however, subdivides it into two 
groups, the “unistelles” and the “ bistelles,” of which the former is a very natural group, 
the modern family Didemnidse. 
This arrangement of the Ascidise Compositse was generally accepted until 1872, 
when Giard published an important memoir,^ in which he gave a classification based 
mainly upon the method of gemmation. He divided the Synascidise (in which he 
included the Social Ascidians of Milne-Edwards) into three great groups — (1) the 
Catenatse, in which the buds are formed at the base or posterior end of the body ; ■ 
(2) the Glomeratse, in which the buds are developed from the region of the repro- 
ductive organs of the parent ; and (3) the Eeticulatse, in which the buds are formed 
as outgrowths from the part of the body where the thorax and abdomen join. 
The group Catenatse includes three families, the Clavelinidse, the Perophoridse, and the 
Botryllidse, but M. Giard gives no sufficient reasons for placing the first two families 
in the Compound Ascidians, and as von Drasche has since pointed out, the third one 
does not really exhibit the essential character of the Catenatse. The second group, the 
Glomeratae, corresponds to Milne-Edwards’ “ Polycliniens,” in addition to his 
“ Didemniens bistelles,” and is divided into two families, the Polyclinidse and the 
Distomidse, both of which are still retained, but have undergone some changes in 
constitution. The “ Didemniens unistelles ” correspond to Giard’s third group, the 
Reticnlatse, and are characterised by gemmation taking place from the region at the 
posterior end of the branchial sac and by exhibiting the phenomenon of embryonic 
blastogenesis. This is a natural enough section, which includes two distinct families, the 
Didemnidse and the Diplosomidse, which Milne-Edwards to a certain extent confused 
with one another. The Glomeratse cannot stand without several changes which have 
since been made by von Drasche, and which really reduce it merely to Milne- 
1 Recherches sur les Ascidies Compos4es ou Synasciclies. 
