PHILIP A. MUNZ 9 



families, in fact, do not form a dichotomy, but the one leads to the 

 other and is connected with it by so many genera that the dividing 

 line is in some cases difficult to find. In a general way the former 

 one is characterized by the presence of several to many antenodals 

 and by cross-veins generally being present in the quadrangle and 

 subquadrangle. M3 separates from M1+2 nearer to the arculus 

 than to the subnodus, and the nodus is far out in the wing, at from 

 one-third to one-half of the length of the wing. These tendencies 

 in venation are all fairly constant and, as will be shown later, in- 

 dicate a rather primitive condition. The Coenagrionidae, on the 

 other hand, have two antenodals generally and no cross veins in 

 the quadrangle or subquadrangle, with but few exceptions. In 

 one line only, that of the Lestinae and Lestoidinae, does M3 retain 

 its original condition of branching off from M1+2 near the arculus, 

 in the other groups its point of origin migrates outward. The 

 nodus, however, is drawn toward the base of the wing and ranges 

 in position from about one-seventh to one-third of the length of 

 the wing. The more detailed account of the exact steps in phy- 

 logeny may better be discussed under the families and subfamilies. 



The Agrionidae 



As was said before, this family includes the old group Calop- 

 teryginae which was divided by de Selys into five legions: Thore, 

 Calopteryx, Euphaea, Lihellago, and Amphipteryx. Venationally 

 these five groups are better arranged in three subfamilies, as was 

 done by Professor Needham (1903a), which I shall call : Polythorinae, 

 to include the legion Thore; Agrioninae, for the legion Calopteryx, 

 and the Epallaginae, for the last three legions. To these must now 

 be added a fourth subfamily, the Disparocyphinae for the genus 

 Disparocypha (fig. 37), recently described by Dr. Ris (1916b). 

 This very remarkable genus will be discussed later with the reasons 

 for its segregation from the others. 



The affinities of the Agrionidae with Epiophlebia (fig. i) are 

 evident enough: the large number of antenodals, the origin of Ms 

 far before the subnodus, and the location of the nodus far out in 

 the wing. 



MEM. AM. ENT. SOC, 3. 



