PHYSIOLOGY OF DIGESTION IN BLATTIDAE 383 
the cushions. Its position seems to indicate that it is the effec- 
tive agent in making the gizzard the strong sphincter organ 
which it is. 
Jordan states that in some insects the posterior, projecting 
part of the gizzard may project completely through the stomach 
when it discharges food, so that the food enters the small in- 
testine and not the stomach. This adaptation results in pro- 
tecting the stomach cells, which have no chitinous wall like 
other parts of the alimentary canal, from actual contact with 
the food. In the American cockroach I find that these condi- 
tions do not hold. Two facts bear out this conclusion: 1) 
longitudinal sections of gizzard and stomach show that the giz- 
zard projects only a short distance into the stomach; even if 
the epithelial folds were flattened out it would fall far short of 
extending through the stomach; 2) when food which can be 
recognized is fed, it can be later found and identified in the 
stomach cavity. Therefore, it is certain that the function of 
the projecting gizzard is not to protect the stomach. 
Ramme’s idea of the function of the projecting gizzard is 
quite the opposite of Jordan’s. He worked on cockroaches 
and other insects and decided that the projecting part does not 
serve to protect the stomach epithelium, but to direct food into 
a region of the stomach well below the coeca, these being di- 
verticula of the extreme anterior end of the stomach. He as- 
serts that the coeca are secretory in function, and that the 
hinder gizzard by directing the food into a region well behind 
them prevents their being clogged with food and so hindered 
in their secretory processes. I shall show that the coeca are 
not special secreting organs. Moreover, the fat I fed did enter 
the coeca in large amount. So the projecting gizzard does not 
serve to protect the coeca in Ramme’s sense. It seems likely 
that the arrangement of parts prevents the hard and sharp 
pieces of food from passing into the comparatively delicate 
coeca. 
Stomach and coeca. Various investigators have discussed the 
question of secretion in the insect stomach, and especially in 
that of the cockroach. The first phase of the question con- 
