456 



In the last day of larval life they proliferate, amitoticalW 

 it seems, and in the larva sixteen hours alter defaecation may 

 extend over a considerable region of the dorsal part of the 

 abdomen. Usually, however, they are confined to two chains, 

 several cells in breadth, on either side of the heart. 



In this condition the cells remain during the early pupal 

 period. Gradually their cytoplasm becomes more homo- 

 geneous, and in the pupa shortly before emergence they may 

 be observed as two irregular chains of unconnected groups of 

 cells, running along the mid-dorsal portion of the abdomen. 

 During pupal life the large cells as they occurred at the end 

 of larval life seem to have undergone a process of incomplete 

 fission, so that one now finds, not chains composed of indi- 

 vidual cells, but chains of small groups of disc-shaped cells, 

 arranged behind one another in little groups representing the 

 cells from which they have been produced (fig. 171). 



Within their clear, heavily eosinophilous cytoplasm lie 

 numerous heavily chromatic granules which usually hide the 

 nucleus. I have observed these glands in the free-living wasp, 

 nine days old, and there can be no doubt that they persist 

 throughout life. 



It seems impossible to regard them as anything but 

 internally secreting glands. Weismann observed certain large 

 cells in close connection with the heart in Diptera, and spoke 

 of them as the ''cell ohaplet." Lowne (1890) observed the 

 same cells, and though he found them in the adult insect, 

 he regarded them nevertheless as young fat cells. I do not 

 know whether they are identical with the structures above 

 referred to; these have, however, a remarkable resemblance 

 in the immature state to fat cells. It is necessary also to point 

 out that they do not constitute the pericardial septum, this 

 structure being absent in the adult Nasonia, 



THE FAT-BODY. 



It is to the great development of the fat-body, together 

 with the disintegration of most of the larval structures, that 

 the pupa of the insect owes its semi-fluid consistency, and the 

 apparent lack of organization that a superficial examination 

 first reveals. 



In the newly hatched larva the fat-body is in the form of 

 a number of large rounded cells, with very faintly granulated 

 protoplasm and a large heavily granular nucleus, lying loose 

 within the haemocoele. A single cell is usually large enough 

 to occupy the greater part of the distance between the intes- 

 tine and the body wall, but sometimes the cells lie ''two deep." 

 Great gaps separate adjacent cells, and through these the blood 



