338 



process of segmentation of the legs lias proceeded far. In 

 the larva ten hours aft-er defaecation a coxa is distinctly- 

 visible behind the femur; the region where the leg showed its 

 origfinal bending^ marks the beginnino^ of the tibia: the tarsus 

 is also clearly visible; on the third tarsus at least four seg- 

 ments have been produced, on the other tarsi joints are just 

 forming. No trochanter is visible yet. 



The wings have meanwhile grown in size. The first wings 

 are now in the form of two great hollow sac-like pockets, on 

 either side of the mesothoracic segments; the hind wings are 

 much smaller. 



In the larva a few hours later the legs have grown so 

 long that they are found beneath the thorax, and their distal 

 ends begin to grow backwards ; the wings continue to grow in 

 length, and likewise become forced backwards. 



The proximal wide ''mouth" of the wings contracts more 

 and more, and the great sac-like structures transform, in the 

 late larva, into others having the shape more nearly of tlie 

 wings of the adult. Rapid growth of the legs continues, so 

 that just before the end of larval life the first leg has grown 

 backwards nearly to the end of the thora^s: ; the second about 

 one-third the distance down the abdomen; the third about 

 one-quarter the length of the abdomen from the end. The 

 wings also, especially the first wings, have become very large 

 and have enveloped the sides of the thorax. 



All the segments of the legs, except the trochanters, 

 are cleary seen ; but the legs themselves are thick fleshy struc- 

 tures, resembling only in a general way the legs of the adult 

 (fig. 16). The same thing has been described above in the 

 mouth-appendages . 



At an early stage in their formation as distinct append- 

 ages, t.e., in the resting larval period, tracheoles began to 

 extend into the wings and legs. Each leg is provided with 

 a single long tracheole occasionally branching into two parts 

 distally. The wings, on the other hand, are well provided 

 with tracheoles. Their actual structure and their history 

 within the wings will be described later; it will suffice to 

 refer here merely to their general disposition within the 

 wings. 



Running along the lower (anterior) border of the wing 

 are a pair of tracheoles (figs. 44, 65), one of which is bifur- 

 cated distally. Tliey appear to communicate in the proximal 

 portion of the wings. Passing down the middle of the wing 

 are a number of tracheoles, which appear also to be branches 

 of a single large tracheole at the base of the wings, this large 

 tracheole giving off a pair of smaller vessels, each of which 

 bifurcates in about the middle of the wing. One of these 



