(34) 



428 0. GOKDON HEWITT. 



extent by those of the other system. They supply chiefly 

 the large muscles of the thorax. Laterally a series of sacs 

 (l.th.s.) extends autero-dorsally in an oblique direction, ex- 

 ternal to the sterno-dorsales muscles to the humeral region. 

 From the first of these sacs a large number of tracheal twigs 

 arise and supply the muscles of the wing and the anterior 

 sterno-dorsales muscles. Ventral to this sac a large sac 

 (m.v.s.) penetrates internally between the anterior and median 

 sterno-dorsales muscles and supplies the lower dorsales 

 muscles. From the dorsal side of the distributing sac a 

 number of sacs arise, some of which penetrate between the 

 sterno-dorsales muscles and supply the upper dorsales mus- 

 cles. A more posterior set supplies the posterior regions of 

 the dorsales muscles, ramifying between them in a very 

 extensive manner, some ultimately terminating in the tracheal 

 sacs beneath the scutum and the scutellar sac (sc.s.). 



The abdominal spiracles differ in number in the two 

 sexes. In the male there are seven pairs of abdominal 

 spiracles ; in the female I have only been able to find five 

 pairs. In both sexes each of the large tergal plates which 

 cover the abdomen has near its lateral margin a small circular 

 spiracle. The first abdominal segment which has fused with 

 the second has a pair of small spiracles (see fig. 8) slightly 

 anterior to those of the second (apparent first) abdominal 

 segment. In addition to these the male possesses two pairs 

 of spiracles in the membrane at the lateral extremities of the 

 rudimentary sixth and seventh abdominal segments (see 

 fig. 5). In the female I have been unable to find any addi- 

 tional spiracles. Each of the abdominal spiracles is provided 

 with a vestibule and atrium which are separated by a valve 

 controlled by a minute chitinous lever. All the spiracles 

 of the abdomen communicate with trachese which ramify 

 among the viscera and fat-body; there are no tracheal sacs 

 in connection with these spiracles. 



