112 



hypothetical type. A comparison of the wings of different 

 species of cockroaches shows that the increase in the num- 

 ber of the branches of the radius has been greater in some 

 than in others ; and a study of the one figured indicates the 

 metliod of increase, which is doubtless as follows : the 

 branches have been added, one after another, to the tip of 

 the trachea, R^ there being a migration of the base of each 

 accessory trachea towards the base of the wing, thus making 

 room for the addition of new branches. In this case the 

 first accessory vein is the proximal one ; but in this particu- 

 lar instance it is impracticable to number the accessory 

 veins, owing to a splitting of some of them which is taking 

 place. 



Accessory veins added proximally. — A study of the cubital 

 trachea of the nymph of the cockroach figured above (Fig. 

 lo, Cu) shows that accessory veins are being developed on 

 the caudal side of vein Cu^. It can also be seen that the 

 distal members of the series of accessory tracheae are well- 

 developed, and that the growth of additional ones is taking 

 place in the disk of the wing at the proximal end of the 

 series. In this case the first accessory vein is the distal one, 

 and the accessory veins should be numbered as indicated in 

 the figure. 



Accessory veins interpolated. — Reference has been made above 

 to a splitting of some of the accessory tracheae developed on 

 the cephalic side of the radial trachea in the cockroach under 

 consideration (Fig. lo). As a vein is developed about each 

 of the divisions of these tracheae, it is evident that at the 

 same time that accessory veins are being added distally to 

 this series others are being interpolated. In cases of this 

 kind it is impracticable to number the members of a series of 

 accessory veins. 



The suppression of the dichotomous branching of 

 veins. — In the more highly specialized of the many-veined 



