100 NEUROPTERA. 
3 Q. Pterostigma of the front wings, *7 ¢,°8-9mm. @ long, surmounting less than one cell (91°/, ¢, 
72:25 °/, 2), onecell (9 °/, 3, 26°/, 2), or more than one (1°75 9/, 9); of the hind wings -85--95 mm. 
long, surmounting less than one cell (91°25 °/, 3, 75°/5 2), one cell (8°75 °/, 3, 23:25 9/, 2), or more 
than one (1°75 °/, 2). 
Antenodal cells on the front wings 4 (73°6 °/, 3, 74:2°/, 2), 3 (15%, 5, 20°3°/, 2), or 34 (113%, o, 
55 °/, 2); on the hind wings 3 (96-25°/, 3, 93°/> 2), 4(25%Jo ds 7%Jo Y), or 2+ (125%, 3). 
Dimensions—Abdomen, 3 21°5-23, 2 20°5-22; hind wing, ¢ 17-19°5, 9 18-20°5 mm. 
Hab. Mexico, Tepic (coll. Calif. Acad. Sci.: 1 3, 1 2), Aguas Calientes (coll. 
Deam: 1 ¢), Guadalajara (Schumann: 2 2 ), Tlalpam (Barrett, colls. Adams, P. P. C.: 
73,4 2), and Mexico city (Barrett, colls. McLachlan, Adams, P. P. C.: 832 3,219). 
Collected at Mexico city and Tlalpam in July and September, at Aguas Calientes 
in December. . 
There does not seem to be a constant dissimilarity in the male appendages of 
var. nahuana and of A. agrioides type. The colour-differences are given in the’ key. 
The varietal name is modified from the Nahua, or Nahuatlan, family of tribes, the 
members of which inhabited much the same region as does this variety. 
“ Division II. * 
The black biserial hairs, spines, or bristles on the legs short (each one usually equal 
- to or shorter than the interval separating it from its next neighbour). Tarsal claws 
each with an inferior tooth. 
The forms which comprise this present division were mostly included in the “ grané 
. genre Agrion” of de Selys (1876). ‘The various “ sous-genres” into which he divided 
it he arranged in two series, characterized solely by the presence or the absence of an 
apical ventral spine on the eighth abdominal segment of the female. Even if it be 
true that this structure is a sure guide to affinities, such a character is very unsatis-. 
factory, from the practical point of view, as a means of identification. Considerable 
evidence, too, may be brought against the view that the presence or absence of this 
spine is of such great importance f. 
The key given below is based on extensive series of statistics gathered in the hope 
of finding characters common to both sexes. Broadly speaking, the genera are arranged 
in an order representing a reduction from a more numerously- to a less numerously- 
veined condition, just as the other legions of Agrionine are arranged in this work. 
The statistics show that it is futile to expect any character in this group to be so 
constant as to afford a sure and unvarying generic diagnosis. Each genus must, in 
consequence, be considered as characterized by the combination of a number of features 
any one of which may be absent in one or other member of that genus. 
* For Division I. see p. 65. 
t See Ent. News, ix. p. 72 (1898), and data given under Ischnura demorsa, I. denticollis, and Leptobasis 
vacillans, infria. As de Selys himself pointed out, Enallagma and Agrion differ only in the presence or absence 
of the spine, yet his system rather emphasizes this slight difference than their apparently close relationship. 
