442 



with great granular nuclei 13jut in diameter, are connected 

 along their length by means of the fine membranous peri- 

 toneum (fig. 146). And as the growing epithelium forces these 

 remains of the old intestine into the lumen, the hepatic caeca 

 are pulled bodily in with them along the whole lengtli of 

 the intestine. In the larva sixteen hours after defaecation 

 the hepatic caeca are being slowly but surely engulfe J (fisfs. 

 147, 148), and six hours later have entirely vanished. To 

 the debris within the alimentary canal is also added the third 

 (posterior) caecum. This becomes drawn into the midgut in 

 a very similar manner; dead larval epithelial cells at its base 

 fail to disintegrate, and the surrounding cells of the new 

 epithelium pushing these cells inwards cause the third caecum 

 to be slowly drawn into the lumen where it disintegrates 

 along with the other disorganized tissues. 



Even before being absorbed the hepatic caeca show signs 

 of degeneration, small globules of cytoplasm being thrown 

 into their lumen. 



The function of the renovated epithelium appears to 

 be to absorb this dehris (fig. 154), perhaps after it has been 

 digested by the enzymes liberated from the disintegrated 

 hepatic caeca. At any rate, a marked absorption of the 

 dehris commences in the fresh pupa. Muscular contractions 

 in this region later drive part of the contents into the posterior 

 portion of tlie gut, and here the apparently indigestible 

 cell membranes of the old larval epithelium accumulate, and 

 may persist in small quantity till the emergence of the wasp, 

 when they are voided through the anus. But the greater 

 part of the fine granular dehris soon disappears. 



Meanwhile in the late hours of larval life a few cells 

 from the rear of the great conical circumoesophageal imagina/1 

 ring have grown in as a short solid mass of cells into the 

 anterior end of the midgut (figs. 153, 154). They push the 

 adjacent intestinal cells along with them and the two co- 

 operate to form a temporary obstruction which prevents the 

 dehris within the intestine (especially when the muscles of 

 the anterior portion contract) from entering the foregut. This 

 ''plug" does not^ however, develop till shortly before pupa- 

 tion, and closure of the anterior region of the midgut during 

 the period which intervenes between defaecation and this, is 

 brought about by a muscular contraction here, which causes 

 considerable folding of the epithelium, and a consequent 

 closure of the passage. The general nature of the regenerated 

 midgut is seen in fig. 154. 



The cells of the newly formed epithelium, which have 

 now attained quite a large size, having performed their func- 

 tion, now begin to disinte2:rate, and for a second time an 



