740 



PROCEEDINGS OF THE NATIONAL MUSEUM. 



VOL. XXVI. 



lines of their own, speciiilizing highly. The shifting of the anal vein 

 in the fore wing, the fusion of the ])ranches of the media be3^ond the 

 arculus, the development of a strong anal loop (of almost Cordulegas- 

 terine form, to be sure), the retraction of the nodus in the hind wing, 

 and especially the general reduction of cross veins and the nan-owing , 

 of the apex of the Aving, mark thii^ as a peculiar group, more distinct 

 than any other Avithin the Libellulid series, and well worthy of sub- 

 family rank. 



lAheUullncV, indudliig ConlaluicV, k. dr. — Passing Macromla, and ' 

 passing also a few isolated forms with triangles in the fore wing four i 

 sided and in the hind wing often little retracted and with the anal loop 

 short, indistinct, or wanting^' — all lateral offshoots, 1 believe, from 

 near the bottom of the Libellulid series — we come upon a series of ■ 

 closelv related forms, the most extensive and flourishing in the order, i 

 It Avere idle to attempt to indicate all the lines of specialization to be i 



Fig. 29.— AViNGS ok Tiirlioptiri/.r hagciii Selys, details in part omitted im fore wing. 



found in the Avings of this series. We Avill endea\^or to point out only 

 a fcAv of the leading tendencies that are superadded to the more gen- 

 eral ones already noticed. 



It is most interesting to note, in this series, hoAV the two venation- 

 evolving processes already discussed, vein adjustment and vein differ- 

 entiation, have alternately held sway. Thus, in the ffrst instance, vein i' 

 adjustment has been carried out to a very moderate extent, vein differ- 



«I pass these by because I have not found the genetic thread that will bind 

 them into a natural series. I haA'e seen but few of them {Tetrathemis, fig. 10; 

 Nannodythemis, Nannothemis (Plate XLIII, fig. 2), Nannodiplax, and Nannophya), 

 and in this paper we are dealing only with lines of development. These are among 

 the rarest of specimens in collections, and differ so much among themselves that 

 there are of them almost as many genera as species. They have all the earmarks of 

 developmental remnants. If Karsch's group Nannophyx (Ent. Nachr., XV, pp. 

 245-26:>) should be made to include all these forms, it should include also Cordnlephya 

 and perhaps fdiiovy.r, which have more affinity with some of these than with any of 

 the CorduHiiic proper. 



