349 



cell division takes place, so that in the larva, at the time it 

 defaecates, the integumental imaginal discs contain about four 

 times the number of cells that we see in the newly hatched 

 larva. Whether these cell divisions occur during the moulting 

 period, or whether they occur gradually throughout larval 

 life I am unable to say. Even this extensive multiplication 

 is not sufficient, however, with absence of actual cell growth, 

 to enable the cells of the imaginal discs to maintain their eariy 

 appearance and at the same time retain their function of 

 forming an unbroken body layer, such as is necessary in the 

 secretion of a new cuticle in the period just preceding a moult. 

 The difficulty is overcome by the cells gradually assuming a 

 curious shape; their outer ends develop into thin flat discs, 

 their inner ends containing the nucleus become long and 

 narrow. Thus, while the outer portions of the imaginal 

 embryonic cells combine to present an unbroken surface — a 

 true pavement epithelium — ^the inner ends, which give the 

 predominating appearance to the disc, are long and narrow 

 and separated by wide spaces. The large specialized larval 

 cells, on the other hand, do not undergo these changes ; on the 

 contrary, they retain their early shape and number, and the 

 only visible change which they undergo is a great increase in 

 size, an increase approximately proportional to the growth of 

 the larva as a whole. 



These large larval cells of the imaginal discs co-operate 

 during the feeding period to form the various cuticles. 

 When newly formed, the cuticle, which is simply a direct 

 secretion from the ectodermal cells, embraces these very 

 closely; gradually it loosens itself, when the integument 

 begins to secrete a second cuticle, inside the first, the two 

 cuticles being very clearly visible in sections through the 

 integument (figs. 47, 55). 



At the time when the larva begins to defaecat-e the 

 ectodermal cells begin to enter upon a period of profound 

 changes. The nuclei have become very large — indeed, their 

 growth appears to have kept pace with that of the whole cell 

 — but the chromatic contents appear curiously disorganized. 

 Each contains a relatively gigantic nucleolus. Then follows a 

 period of cytoplasmic disintegration. The entire cell contents 

 break up into numerous minute globules (figs. 28, 29) which 

 have a curious resemblance to nucleated cells, each consisting 

 of a clear outer zone, and containing a heavily staining body. 

 They are, however, purely disintegration products of the 

 large cells and are not to be confused with the leucocytes, 

 which are much larger than these globules; their^ curious 

 construction is probably due to some obscure physical con- 

 dition of the disintegration products of the cell. These 



