204 Annals Entomological Society of America [\'o\. III. 



In ho])Cs of beinjj; able to study to Ijetter advantage the type 

 of secretion and of better understanding the relation of its 

 discharge to the intake of food, I undertook a series of feeding 

 experiments. Larvae were placed in small glass vessels with 

 clean tap-water and kept there without food for periods 

 varying from one to twelve days. The larger larvae invariably 

 jnipatcd before the experiments could be completed, so it was 

 necessary to use larvae which were less than half grown. Larv^ae 

 that had fasted for varying ])eriods up to twelve days were 

 killed and sectioned for determining what changes, if any, the 

 mid-intestinal epithelium underwent during the fasts. Like- 

 wise larvae starved for two, three, four, seven and twelve days 

 were removed to vessels containing abundance of food, and then 

 after feeding for periods varying from half an hour up to seven 

 hours, they w^ere killed and sectioned for studying the epithelial 

 changes accompanying feeding. 



From a study of the preparations made from these larvae 

 it was found that where the fast was prolonged more than four 

 days, the mid-intestinal epithelium became so reduced and 

 weakened that it failed to recover even though the larva sur- 

 vived the fast and fed freely for a time (Fig. 40). Simpson ( '01, 

 in an unpublished paper) found a similar degeneration of the 

 epithelial cells of the mid-intestine of Clisiocampa larvae that 

 had fasted for five days. When the fast is prolonged four days 

 the cells take on a distinct vacuolate appearance associated 

 with a heavy discharge of secretion (Fig. 39). After six or 

 seven days this vacuolate condition has disappeared and the 

 cells become greatly reduced in height. They almost cease to 

 excrete digestive fluids and the striate border is then more 

 conspicuous. Aside from a continued reduction in the height of 

 the epithelium, and a greater shriveling of the cells, no further 

 marked changes are brought about, the cells finally dying. At 

 no time, even after a twelve day fast, is the mid-intestine 

 completely free from food material. As a consequence there 

 is a constant stimulus to the excretion of the digestive secretion 

 and as the fast is prolonged the finely granular secretion is 

 poured out and soon comes to fill most of the space between the 

 reduced epithelium and the food content of the canal (Fig. 39, 

 d. sec.) In i)lace of the secretion being stored up in the cells 

 as in the nymjjhs of dragonflies, it is continually poured into the 

 lumen. This process soon drains the cells of their vitality, 



