PHYSIOLOGY OF DIGESTION IN BLATTIDAE 389 
ewsky and Metalnikoff have used the method of pigment feed- 
ing, and conclude that almost all of the absorption occurs in the 
stomach. Metalnikoff fed food containing an iron salt; while 
it was being absorbed he precipitated the iron as Prussian blue; 
he found absorption in the stomach only. 
Plateau, Petrunkevitch, and others, after thorough investi- 
gation, find the crop and important organ of digestion and the 
stomach of less significance. I have verified this, and further 
credence is given to this view by Sayce and others demonstrat- 
ing that osmosis can occur through a chitinous intima. 
My own investigations support those of Petrunkevitch. Fat 
is absorbed by the stomach cells and may be demonstrated in 
them by staining the tiny globules black by osmie acid. These 
globules first appear at very varying times after food is eaten, 
but no doubt only a comparatively short time after the food actu- 
ally enters the stomach, provided secretion globules are present. 
Experiments with colored food have shown that fat may pass 
from the crop very soon into the stomach or may be entirely. 
held back two or three days, even in animals which ‘have been 
previously starved a long time. If the stomach is alone capable 
of absorption, as so many authors believe, it would indeed be 
hardly believable that an almost starved cockroach should be 
denied the power of absorbing or utilizing its food until several 
days after it has devoured it. What really happens is that the 
food begins to be absorbed and utilized by the crop soon after 
it is eaten. ‘ 
The cells of the stomach are differentiated into older cells and 
regenerating ones; only the older cells can absorb. These usually 
occupy only a limited part of the surface of the epithelium, per- 
haps a third or less. The contrast between the absorbing cells 
and the others is striking, as shown in figure 21, where fat is 
being actively absorbed. Occasionally the upper ends of the 
neighboring cells may absorb a little fat. As this is only seen in 
the ends of the cells toward the lumen, it seems that fat is soon 
drawn from the neighboring cells through the lateral cell walls 
into the true absorbing cells. The cells which absorb most 
strongly secrete most strongly, as a rule, and there is almost no 
