62 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 
stigmata are present, not in the lateral walls of the 
pharynx,—in which case they might lead directly to the 
exterior, but in a fold of the ventral wall,—an arrangement 
obviously opposed to any direct communication with the 
exterior. Moreover, he states that the presence of the 
stigmata in the pharyngeal wall of Mossia gives the whole 
structure a general resemblance to the respiratory dia- 
phragm of Doliolum; and, since the stigmata in the so- 
called diaphragm of Doliolum connect the pharyngeal with 
the cloacal cavity, it is clear that Moss believed that the 
‘“‘stigmata’’ in his species were also openings from the 
pharynx into a water-space around it. That they are 
not openings into a true cloaca, like that of Doliolum, is 
manifest from the fact that the intestine in Mossia opens 
directly to the exterior. We are therefore bound to infer 
that the stigmata, if really present, lead into a pair of 
exhalent or atrial canals as they do in larval Ascidians. 
And if Moss was mistaken as to the presence of stigmata, 
there is still less reason for doubting the existence of 
structures homologous with the branchial canals of other 
Appendicularians. They must have been overlooked 
owing to some unusual and deceptive conformation. 
Such a conformation is actually presented by Kowa- 
levskia tenuis (fig. 8, AW). In this species the exhalent 
canals are not long and tubular as they are in the greater 
number of Appendicularians; they are hardly anything 
else than large apertures in the sides of the pharyngeal 
region of the body: their diameter is enormous, and their 
length (from pharynx to body-wall) is reduced to a mini- 
mum. If we bear in mind that—to quote Fol’s words— 
“this little creature is of an extreme delicacy and tran- 
sparency,” and that Surgeon Moss made his observations 
under the ‘‘ peculiar difficulties of study at sea,” we can 
not greatly wonder that he failed to notice spaces so large 
