GENERAL STRUCTURE. 35 



removed, but the fact is evident. From the invagina- 

 tion aforesaid, it follows that the layer which is 

 outside in the skeleton is the inner, or lining one, in 

 the tracheae, while the hypodermis, which originates 

 the chitinous coat (as has already been stated), and 

 lies, of course, beneath it, has its representative out- 

 side the breathing tubes. The spiral thread is pro- 

 duced by the lining membrane, or internal cuticle, 

 forming a chitinous thickening, in a spiral line, which 

 is never continuous for more than four or five turns. 

 Just before one thread terminates a new one starts, 

 to be in like manner followed by another. The tubes 

 are only capable of slight extension, and, when 

 unduly stretched, the membrane ruptures, and the 

 spiral is drawn out singly (as at d, Fig. 6), or a 

 band of four or five threads will separate for a few 

 turns (as at b). The slenderness of the smallest of 

 these tubes, which have neither interior cuticle nor 

 spiral thread, is as remarkable as their number, and 

 the microscope, even at its best, is barely able to 

 trace out their terminations. Of such, a bundle 

 containing a quarter of a million, would scarcely 

 exceed in bulk an ordinary human hair. 



In bees, as in all actively flying insects, the 

 tracheae are accompanied by large air sacs (a, 

 Fig. 6), which are developed in the same manner as 

 the tubes themselves, but carrying scattered venations 

 instead of spiral thickenings of the membrane. In 

 the larval state of comparative inactivity no aerial 

 sacs exist, but they are brought into being during 

 the chrysalis changes. These air sacs have much to 

 do with flight, in a way to be explained when we treat 



E 2 



