/ 483 



notj^en able to^ observe them directly, it would seem that 

 the fibres should be of the T-shaped type. 



Meanwhile the two antennal ganglia have been increasing 

 in size. The cells in the early pupa send out processes along 

 the degenerate antennal nerve tract of the larva, and as the 

 latter is gradually absorbed the fibres of the former develop 

 at its expense. The fibres from the antennal nerve pass 

 inwards into the antennal lobe, and within it meet and 

 evidently form synapses with other fibres given off from cells 

 in the more dorsal parts of the antennal lobes, these fibres in 

 turn passing backwards and upwards as a short, thick, nerve 

 tract which enters the great irregular mass of nerve fibres in 

 the middle of the brain (fig. 234). This ''white matter" 

 of the antennal ganglion is a very large and rounded mass 

 of fibres showing shallow clefs on its surface. 



Meanwhile the great mass of paler cells described above 

 has continued to grow; the cells encroach more and more 

 upon the great degenerate nerve strands of the larva at the 

 base of the two hemispheres in the region between the antennal 

 and ocellar ganglia (fig. 232x) ; and shortly after the first day 

 of pupal life, continuing to absorb the whole larval nerve 

 strand without proliferating, so far as I could observe, any 

 more, gradually replace this, and as nerve fibres from these 

 and other cells lying more on the periphery grow into the 

 dead nerve strand, this is finally absorbed and replaced by 

 the fibres from the imaginal cells- These fibres seem to com- 

 municate with others formed from the inner optic ganglion 

 and the resulting structure is the ''opticon," as Hickson has 

 called it in Calliphora. 



By this means, then, the larval brain is gradually trans- 

 formed into that of the adult. Phagocytes play no part in 

 the process of absorption, but the dead cells serve directly as 

 food for the growing imaginal cells. And although the pre- 

 sence of mitotic figures within the brain is the only clear sign 

 that development is going on at all, yet when a more careful 

 study is made it is soon seen that the brain undergoes as 

 profound a metamorphosis as does any other organ of the 

 body. 



Meanwhile the first ventral ganglion has gradually become 

 incorporated into the brain. In the fresh pupa, although 

 "rejuvenation" of the ganglion has taken place, like most of 

 the other ganglia of the ventral chain-, migration has not yet 

 commenced. But shortly after pupation the cells oi the 

 ganglion begin to migrate upwards along the circumoesopha- 

 geal connectives (fig. 232). In the twenty-six hour pupa they 

 have definitely become a part of the now very complex brain 

 fig. 233), and during the next day they begin to consolidate 



