351 



directions; the outer, most actively migrating cells often 

 show amoeboid processes (fig. 29), and it may be by this 

 method of locomotion that the cells advance. Sometimes they 

 grow right over the dead larval cells (fig. 28) ; at other times 

 they seem to be unable to cross them, and have to await the 

 complete destruction of the larval integumental cells before 

 they can advance further (fig. 32). 



During the last hours of larval life, therefore, the integu- 

 ment consists of areas of proliferation, the cells of which, 

 growing outwards, are actively engaged in replacing the dis- 

 integrated larval cells, or awaiting the total destruction of 

 these. 



Eventually the imaginal discs of the integument meet, 

 a cuticle is secreted and the larval moults, disclosing the pupa. 



At times leucocytes, having disposed of the remains of 

 the larval tissues, are seen crammed with larval debris lying 

 among the proliferating imaginal cells, and evidently provid- 

 ing, by their disintegration, nourishment for the surrounding 

 cells (fig. 31). Sometimes, also, numerous of the large hyaline 

 degeneration globules are seen in similar situations (fig. 31). 



The integumental cells, unable to extend further, now 

 begin to undergo structural changes; at first spindle-shaped 

 (fig. 30), they soon begin to change their general form; in 

 some parts of the integument, especially that of the abdomen, 

 the cells are small and cubical, their outer surfaces very 

 regular; the chitin secreted from them in tliis region is quite 

 smooth. Along the dorsolateral regions of the pupa the cells 

 are generally rather elongate. In the antero-dorsal region 

 of the thorax (especially in the region of the future pronotum) 

 this condition is especially clearly seen. Here the outer ends 

 of the cells develop broad swellings, giving the cells a hammer- 

 like appearance; from these swellings the thick chitin in this 

 region becomes secreted. A somewhat similar condition is 

 seen right at the posterior end of the abdomen. The ectoderm 

 of the propodeum and metathorax is especially remarkable, 

 being in the form of a great accumulation of ectodermal cells, 

 several layers deep, all crushed together, and thus accounting 

 for the contraction which external features show has gone 

 on in this region. 



The secretion of cuticle now goes on very rapidly, and 

 about four hours after pupation, forms a layer nearly as 

 thick as that of cells which are secreting it. The cells from 

 which the cuticle is being secreted, moreover, do not, as a 

 rule, present a perfectly smooth surface, but become so dis- 

 posed as to form a mould on which the cuticle of the imago 

 can shape itself, and the various depressions and bosses, and 

 other sculpturing with which the imago is ornamented, as well 



