380 ; ELDON W. SANFORD 
region in the species I used, and have uniformly found no evi- 
dence of such passages. It may be that in some Orthoptera 
the coeca are special organs of secretion, so that a more direct 
passage would be advantageous. In the cockroach the coeca 
show the same secretory power as the stomach; a direct passage 
for their secretion would, therefore, have no significance. 
The cushions of Miall and Denny deserve special attention. 
They are six in number and situated directly behind the pri- 
mary teeth (figs. 7 and 10). Their surface is provided with 
needle-like spines, and these are especially significant on the 
surfaces directly toward the lumen. <A definite band of striated 
muscle runs from this surface outward and upward to be firmly 
attached to the thick chitinous intima of the anterior surface 
of the teeth (fig. 7). Wilde has described the cushions as nor- 
mally closed, their opposed surfaces fitting to form a tight 
joint. He considered that the tight closure of the cushions 
prevented the passage of food until it is well ground by the 
teeth. Basch thought that the cushions not only pushed against 
each other, but also pushed up closely under the teeth. Ramme 
has also described a close joint by the pressing together of the 
cushions by action of the surrounding ring muscles. Petrunke- 
vitch has described the needles of the exposed surfaces as being 
attached inwardly to muscle fibrils, one fibril running to each 
needle. He believed that the contraction of the muscles caused 
movement of the needles by virtue of the attachment of the 
muscle fibrils to the bases of the needles. Ramme has ridiculed 
this view; he finds the muscles attached to the intima, not to 
the spines. 
I have made many series of transverse and longitudinal sec- 
tions of the gizzard, and never have I found the cushions closely 
appressed, and only in one case were they near each other. 
The cushions appear well separated from each other, as I have 
figured in transverse and longitudinal sections (figs. 7, 10, 15). 
As my preparations represented every stage of feeding and star- 
vation, we may safely conclude that in the American cockroach 
the close opposition of the cushions occurs rarely or never and 
has no significance in the retention of food. The cavity of this 
