THE DEVELOPMENT OF SYMBRANCHUS MARMORATUS. 27 
being just a simple ring-shaped opening, but there is a long, 
very much coiled intestine. If there is a similar absence of a 
specialised valve combined with the presence of a very long 
intestine in the other members of the above-mentioned orders, 
then it would seem as though the absence of pyloric caeca and 
any specialisation of the pyloric valve in these Teleosteans 
were compensated for by the presence of a very long intestine. 
A comparison of such forms as Cottidae with one of the 
Catosteomi (Stickleback) is instructive as illustrating how the 
degeneration of pyloric caeca (according to Goodrich (3) pyloric 
caeca are lost, as a rule, in this suborder) may possibly bring 
about an appearance such as that described in Symbran- 
chus. In Cot t us the short finger-like processes lying round 
the front part of the duodenum, with their blind ends lying 
on the stomach, form a circle of pyloric caeca. A dissection 
of this part of the alimentary canal reveals a rosette-like 
valve (the protrusion of the stomach into the duodenum) 
surrounded by the individual openings of the blind sacs. 
At first sight there is no constriction visible between the 
stomach and intestine in the stickleback (Gaste rosteus 
aculeatus), but the constriction is really present though 
hidden by two broad-mouthed blunt bulgings of the 
duodenum in the direction of the stomach. One of these 
pockets is bigger than the other as a rule (in one specimen 
examined the smaller of the two was very small). Both are 
well bound down to the alimentary canal by connective tissue — 
in fact, they seem to show a tendency to fuse with it. The 
valve is of the same type as that of Sym branch us, so that, 
should the pocket-like outpushings become fused with the 
alimentary canal, then the likeness between the pyloric 
arrangements of the stickleback and those of Symb ranch us 
would be perfect. 
In this case it would seem that, if the pyloric caeca are 
really disappearing in the Catosteomi, then the Sym- 
b ranch us pyloric valve is the result of degeneration of 
pyloric caeca — an indication of specialisation in Sym- 
branchus . 
