361 



the mouth. Their development during the larval and early 

 pupal periods is quite similar to that of the thoracic append- 

 ages, i.e., the cells proliferate, grow outwards, forming a 

 cavity behind them as they do so, and drag the underlying 

 mesoderm after them. The general features of their develop- 

 ment have already been described. 



The mandibular imaginal discs (fig. 47) are particularly 

 interesting. Each consists of a sac of ectodermal cells (lined, 

 of course, with mesoderm), and invaginated well into the 

 cavity of the head. The floor of the ''sac" is many cells 

 thick, the cells themselves being rather smaller than the larval 

 integumentary colls. On the "floor" of the invagination is 

 a small number (about nine) of very remarkable cells; they 

 are club-like in shape, and each bears a long cytoplasmic 

 extension outwards. These processes are so arranged that 

 they possess, together, the shape of the larval jaw, and it is 

 from the termination of these remarkable cells that the 

 minute stylet-like mandible of the larva is secreted. This is 

 seen" clearly'ln fig. 47, which is taken from a larva shortly 

 before entering the second instar. The functional jaw is no 

 longer in communication with the cells which have secreted 

 it; these cells, on the contrary, are now secreting a second 

 mandible, within the first, and the latter will be cast off at 

 the larval moult. 



We see, then, that the larval mandible is formed from the 

 same set of cells which produce the mandible of the imago. 

 The jaw of the larva must therefore be regarded as homo- 

 logous, in part, with the jaw of the mature insect. It is, of 

 course, quite conceivable that this might not have been the 

 case ; however, the fact that it is so can lead to important con- 

 clusions, which will be considered in the second part of this 

 paper. The same thing will be seen later, in connection with 

 the antennae. 



During the feeding period of the larva the mandibular 

 imaginal disc grows in size ; at about the middle of this period 

 the disc has become partly everted, and the projecting portion 

 has the shape of the jaw of the last instar. On account of 

 the much greater size of the jaw at this stage, it is necessary 

 for a much larger part than hitherto of the imaginal disc to 

 take part in its formation, and this is the reason for its pre- 

 cocious evagination. 



During the resting stage the mandibular disc grows 

 rapidly, the cells dividing mitotically. So far as I could 

 observe, the long club-like cells of the disc in its first instar 

 become modified during larval life into cells which do not 

 differ visibly from the others of the madibular disc, i.e., cells 



