178 ME. E. M'lACHLAjS' 02s THE NEFROPTEEA OP 



American Lilellula vesiculosa, Fab., aud other American insects 

 agree therewith in points of generic structure; but the Old- "World 

 species associated with them are divergent in several points (yet 

 agree in having the base of the abdomen inflated in lotli sexes) ; 

 and on a minute revision (a thing hoped for !) of the subfamily 

 Libellulina they would have to be sepai'ated from their American 

 allies. Let us now turn to a familiar group of Old- World forms, 

 of which the British L. cancellata, L., and ccerulescens, Tab., may 

 be considered typical, — forms for which Newman, in 1833, pro- 

 posed tlie generic term Ortlieimm {=Libella, Brauer, 1868, a 

 name otherwise inadmissible, because Selys had previously used it, 

 and in Odonata) . In the more typical forms of Orthetrim, neither 

 6 nor § has the base of the abdomen conspicuously dilated 

 in comparison with the rest. But when we come to Libellula 

 J«rZ>«rc, Selys, (and some others,) we are concerned with an iusect 

 the very adult male of which, in a thoroughly blue-pulverulent 

 condition, is diflB.cult to separate from L. trinacria in a similar 

 condition, but in which the $ has the base of the abdomen in no 

 way dilated. Therefore I should be disposed to consider L, har- 

 lara an Ortlietrum, and to place L, trinacria (and immediate 

 allies) in a distinct group (on account of the female characters), 

 more related to Orthetrum than to LeptJiemis typically. 



Now, as to the sjyecific question involved in the Canarian i. 

 chrysostigma. The materials examined by me are as follows : — 



(1) The mutilated $ type of Olympia, Brulle (remarkable for 

 the presence of .only two rows of discoidal areoles in the anterior 

 wings for the greater part, an accidental condition, of which 

 traces are frequently discernible in other species of the group) ; 



(2) the semiadult S and very immature $ in the collection of 

 the British Museum ; (3) the very adult S captured by Eaton, in 

 my collection. The two females and the semiadult S prove in- 

 contestably that chrysostigma is not conspecific with trinacria. 

 But they show very close connection with harhara, yet are larger 

 (about as large as small examples of trinacria) ; and, at this 

 moment, the size is almost the only tangible point of difference 

 apparent ; for I fail to appreciate any striking difi"erences in the 

 genitalia of the second segment in the c?, or in the vulvar scale 

 (insufficiently examined on account of condition) of the $ . Prom 

 trinacria the immature or semiadult S of chrysostigma (and also 

 of larhara) may be at once separated by the thoracic markings 

 (combined with the much smaller pterostigma) ; from highly adult 



