136 THE WINGS OF ISOPTERA 



The descriptions by Desneux of this part of the wing are incomplete; 

 that of Silvestri is obviously incorrect, as he designates a vein that lies a 

 considerable distance from the costal margin of the wing as the costa; 

 Holmgren, whose account is the most complete, has been misled, I believe, 

 by an aberration in the tracheation of the wing studied by him, which led 

 him to believe that vein Ri, which he terms "radius," is represented by 

 two veins. 



Recently, through the kindness of Mr. Walter W. Froggatt, who first 

 described this remarkable insect, I have received a specimen of it. In the 

 fore wings of this individual the tracheae are clearly visible, and in the right 

 wing the aiTangement of the principal branches closely resembles that of 

 the hypothetical type. Figure 128 represents the base of this wing; the 

 entire wing is figured later. Great care has been taken to represent 

 accurately the courses of the tracheae in that part of the wing figured here, 

 the figure being based on a photomicrograph. 



In the left wing of this individual the branching of the tracheae is some- 

 what different. It seems to me that the tracheation of the right wing 

 should be regarded as typical, as it corresponds closely to that of the 

 hypothetical primitive type, and that of the left wing as aberrant. 



Regarding the identity of the radial trachea (Fig. 128, R) there can be 

 no doubt; its position and manner of branching are typical, except that the 

 branching takes place unusually near the base of the wing. Trachea Ri 

 branches off from trachea R near the end of the first third of the length of 

 the scapular shield. A short distance beyond the point where trachea Ri 

 separates from trachea R, the trachea of the radial sector divides; the 

 branch R2+3 is much smaller than the branch R4+5. Trachea R2+3 

 divides a short distance beyond its point of origin, near the middle of the 

 length of the scapular shield, one branch extending into vein R2 and the 

 other into vein R3. Trachea R4+5 passes into vein R4+5, which first 

 divides beyond the end of the scapular shield. 



Immediately in front of the radial trachea is the subcostal trachea (Fig. 

 128, Sc). This enters the wing separate from the radial trachea, but the 

 two tracheae extend closely parallel in the first third of the length of the 

 scapular shield; the subcostal trachea then extends obliquely towards the 

 costal margin of the wing; near the beginning of the last third of the 

 length of the scapular shield, the subcostal trachea forks, trachea Sc^ enters 

 vein Sc2, which is a well-developed vein. Vein Sci is very indistinct. 



The costal margin of the wing is not greatly thickened and no costal 

 trachea is visible in this specimen. 



The medial trachea (Fig. 128, m) enters the wing separately; it is 

 closely parallel with the radial trachea in the basal part of the scapular 

 shield; it crosses the median furrow and enters the strong media near the 



