398 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
en- 
^at 
";i1-V]i1r;.-'!ft7 " 
oe- 
mm 
-fp 
-sd 
tend close to the endostyle, which is 
narrow and inconspicuous. The stig- 
mata at both ends of the rows are 
small, but reach their full length at a 
very short distance from the endostyle, 
while at the dorsal ends of the rows 
only the first two stigmata are generally 
perceptibly shortened. In the first or 
anterior series, however, the stigmata 
increase gradually in length to the 
middle point of the series (see text-fig. 
21), and this is also true to some 
extent of the last series. The trans- 
verse vessels are not strongly muscular. 
The dorsal languets are short and 
hooked, and directed posteriorly, the 
recurved point bending toward the 
ventral side. They appear to have 
absolutely no connection with the me- 
dian dorsal vessel, for they are placed 
about opposite the sixth or seventh 
stigma on the left side of the body, 
arising from the transverse vessels of 
the branchial sac. There are no lan- 
guets on the right side. 
The intestinal loop is little or not 
at all twisted in most of the well ex- 
panded zooids, but in other cases it is 
(as usual in the genus Amaroucium) 
twisted through a full half turn, bring- 
ing the stomach, which is oval and 
longer than broad, to the dorsal side. 
It has been suggested by Bjerkan 
(1905) that such twisting, which oc- 
curs in most compound ascidians, may 
be largely due to the contraction of 
the body walls. It is easy to see that 
the twisting might admit of a more 
compact disposal of the parts of the 
Text-fig. 21. — MacrocUnum pomum 
(Sars). Zooirl. X 15. 
