PHYSIOLOGY OF DIGESTION IN BLATTIDAE 395 
venting entrance of any other substance from without; air in the 
remainder of the tubes at normal pressure, so there is no danger 
of sucking action anywhere even if tubes are injured or cut; red, 
if any, in the tubes, indicating a normal content of red fat in the 
tracheae due to metabolic processes and by no chance to the 
conditions of the experiment. The actual set of experiments is 
the following: 
One hundred and twenty-six cockroaches were fed a paste con- 
taining powdered sugar, olive oil, and Sudan III, these were re- 
moved at intervals. Eight hours after feeding three of them 
were removed from the jar, their necks were tightly ligated with 
silk to prevent any fluid used in the experiment from being 
swallowed, and they were immersed in the green mixture of gly- 
cerin and aleohol. After a short period of extreme activity, the 
animals died by suffocation, due to the liquid entering the 
breathing tubes; they were then removed from the green solution, 
washed, and dissected. It was found that the green had pene- 
trated a considerable distance into the tracheal trunks, it had 
been drawn in by the sucking force present in the tubes before 
the animal died. The presence of this green fluid in the tracheae 
during the dissection prevented the entrance of any other sub- 
stance into the tubes, not only by way of the spiracles, but at 
any place, for pressures were equalized. The trunks were cut 
in the region containing the green and the alimentary canal was 
carefully removed without injuring the wall of the crop. It was 
mounted in toto in glycerin and examined microscopically for 
the presence of any red substance in the tubes; such red substance 
would if present mean that red fat had been derived normally 
from the food in the crop’s lumen. But no red fat could be 
detected in the tubes. 
Three more of the animals were removed and treated similarly 
sixteen hours after feeding, three more at twenty-four hours, and 
three more at the end of eight eight-hour period throughout two 
weeks. In none of the 126 animal used could distinct redness be 
observed in the tracheae. Other animals were fed and dissected 
under the same conditions except for the fact that the crop was 
slit longitudinally during dissection; these showed no red sub- 
