742 PROCEEDINGS OF THE NATIONAL MUSEUM. 



margin of the wing, the hind angle only moderately enlarged, special- 

 ization by reduction seems again to have dominated, producing a group ij, 

 of genera of which Macr(jthendf< is a good exponent (Plate XLVI, ' 



fig.l). 



Then there is the important group of genera allied to LiheJhda, in 

 which the branches of the media at their departure from the arculus 

 are hardly fused, in which the narrowing and transverse elongation of 

 the triangle and the deflection of the anal vein before the triangle have i 

 progressed very far, but in which only a very moderate reduction of 

 cross veins has occurred (Plate XLVIII). The tendency of vein if^ ' 

 to become undulate is also a feature of this group. 



There is also the heterogeneous group of genera allied to Neurotliemu. 

 These will at least agree in exhibiting the highest degree of vein r| 

 adjustment together with the lowest degree of vein reduction to be ([ 

 •found in the series (lig. 18).' 1 



At some point in this series not remote from Celithemk (Plate i 

 XLVI, fig. 2) may have set in the extensive reduction of cross veins 

 characteristic of a verj^ large number of genera allied to Symjyetnim. 

 (See Plate XLVI, fig. 3, and Plate XLVII, fig. 1.) 



Another offshoot from near the same place, combining in a high 

 degree both kinds of specialization, culminates in Tramea and its 

 allies (Plate XLIX), which I regard the most specialized of Libel- 

 lulinjv, if not of all Odonata. 



In the three last-mentioned groups there is a pretty adjustment of , 

 the second cross vein in the space between veins J/j and J[/^ in oppo- ij 

 sition to the proximal end of the radial supplement. Plates XLV and 

 XLVI and fig. 2 of Plate XLVII will show its increasing definiteness 

 and obliquity of position and the perfecting of the brace of which it 

 is a third part, the other parts being the supplement and the oblique i; 

 vein. A trivial character this ? Indeed it is but a straw, yet it shows 

 the way biologic winds blow. 



ZYGOPTERA. 



In this suborder we have again two families, the more generalized 

 abounding, as before, in venational experiments (if I may so speak), 

 some of which have been abandoned in recent times, the more special- 

 ized, being more homogeneous, characterized by a single principal 

 trend in its wing development. The suborder as a whole has retained 

 in recent times two primitive characters, in the similarity of fore and 

 hind wings" and in the total lack of supplements. 



A. CalojAerygidx. — There are four fairly distinct types of venation 

 in this family, the most peculiar of which is represented by the fossil 

 genus hophlebia (fig. 31). In this type the hind wing was as much 

 dilated at the base as in the Anisoptera, but by very different means. 



" For this reason we figure in most cases but one wing. 



