12 



BULLETIN 841, U. S. DEPARTMENT OF AGRICULTURE. 



frequently found several inches above the lower end of the boring, 

 moving through the solidly packed "sawdust." As the larva ap- 

 proaches maturity it is doubtful if it ventures into the upper and 

 slender part of the stem, but it still reworks the frass farther down, 

 enlarging the bore in places. 



The number of ins tars is difficult to determine, owing to the larval 

 practice, just referred to, of passing all the frass several times through 

 the digestive apparatus. Nearly all the cast skins disappear com- 

 pletely under this treatment, only the heavily chitinizod parts such 

 as the jaws and caudal horn being recognizable in the burrow. Care- 



Fig. 7. — Western grass-stem sawfly: Mature larvas removed from their galleries. Enlarged 4 diameters. 



ful investigation of these fragmentary remains appears to establish 

 the fact that there are four molts. The contents of innumerable 

 stems have been examined with scrupulous care and with varying 

 results. In a few cases as many as four sets of jaws and in others 

 four caudal horns have been found, mixed with the frass within the 

 stems. Seldom were more than four sets removed from a single 

 stem; usually only three were found. As is stated elsewhere in this 

 paper, it is no uncommon tiling to discover two and even three 

 larvse mining a single stem, although but a single individual can 

 possibly reach maturity with the amount of nutriment contained in 



