PHYSIOLOGY OF DIGESTION IN BLATTIDAE 367 
Salivary glands. Various investigators have found that the 
reaction of the saliva in Orthoptera is never acid, but may be 
neutral, as Plateau found in Blatta orientalis, or alkaline, as in 
Blattella germanica. Jordan states that the saliva has about the 
same digestive power on carbohydrates that human saliva has. 
Crop. If we consider the function of the crop, there are two 
principal possibilities, 1) secretion and 2) absorption. <A third 
possibility is storage of food, and this is no doubt an important 
factor, enabling the cockroach to store here much of a very large 
meal and utilize it during a foodless period of several days or 
weeks, especially during bad weather in its native home in the 
tropics. The large size of the crop is evidently an important 
factor in making them as hardy and prolific as they are. This 
is in accord with Jordan’s theory that the manner of life brings 
about more variation in the foregut of insects than in the mid and 
hind guts. 
I have almost never found the crop entirely empty, though 
after three weeks of starvation in a clean glass jar it may con- 
tain nothing but a small amount of fluid. Under such condi- 
tions, they ingest various non-nutritious substances, such as 
wood shavings and tarsi and pieces of antennae of their mates. 
The content usually appears as a sticky gray mass. After they 
had eaten all they could of the red mass I fed them, their crops 
were very much distended, completely filling the thoracic cavity 
and more than half of the abdominal cavity, extending as far 
back as the fifth abdominal segment. The pressure caused the 
epithelium to bulge out between the muscle strands. Frequently 
this pressure was so great that a part of the food was regurgitated. 
Such a meal may last the cockroach two months or somewhat 
less, as found by dissection at various intervals. This surely 
shows that the crop is important as a storage organ. 
The epithelial cells are all alike histologically and show no 
special characters of secretion or absorption. Experiments to be 
described herein showed that secretion occurred in these cells, but 
methods of direct observation proved insufficient to determine the 
reaction of this secretion. The reaction of the content of the crop 
depends largely on the kind of food in the lumen, partly on the 
