1913.] NATURAL SCIENCES OF PHILADELPHIA. 263 



its contemporary Lithagrion, as may be seen bj- a comparison of 

 those numbered 2, (4?), 12, 156, 17a and h, 19, 20, 21a and h, and 22, 

 on pp. 227 and 232, anted. 



Phenacolestes is more primitive than its contemporary Melanagrion 

 in characters Nos. 1, 2, 12, 156, 17a and 6, 19, 20, 216 and 22, but less 

 primitive in characters Nos. 3, (4?), and 11. See pp. 227 and 231, 

 anted. 



Phenacolestes may have been the ancestor of the New World genera 

 Megapodagrion, Allopodagrion, and PerUestes and of the African 

 Chlorolestes, unless it shall be shown that the less oblique pterostigma 

 possessed by all four of these living genera cannot have been acquired 

 from the form which existed in Phenacolestes. Of these four, Mega- 

 podagrion is most like Phenacolestes. But Megapodagrion and the 

 other three genera may equally well have descended from Melana- 

 grion, with which they agree more closely in the shape of the ptero- 

 stigma, although Melanagrion is figured as having no anal cross-veins, 

 a not very weighty objection in this connection. 



Phenacolestes might be considered ancestral to Heteropodagrion 

 and Heteragrion, were it not that these two have, in at least some 

 species, a greater number of postnodals. 



Phenacolestes was apparently- not the ancestor of: 



Neuragrion, which has more postnodals, and the cubito-anal cross- 

 vein in a different position, i.e., proximal to the level of the first 

 antenodal ; 



Mesagrion, for the second reason just given for Neuragrion; 



Thaumatoneura, Paraphlehia, or Dimeragrion, which have more 

 than one cubito-anal cross-vein and denser venation in the wing 

 generally {Thaumatoneura), or between Cui and Cu2 {Paraphlehia), 

 or as shown in the greater numbers of postnodals (all three) ; 



Philogenia, which has more postnodals and rather denser venation 

 between Mia and M2; 



Rhipidolestes, which has Ms separating from M1+2 more proximad, 

 Rs beginning proximad to subnodus, and more postnodals; 



Podopterijx, which has a denser venation between Ma and M4 and 

 posterior to Cu2, and more postnodals; 



Argiolestes, which has a denser venation between Ms and M4. 



Comparisons with genera of the Old World having more reduced 

 venations are useless, as the equally possible descent from other 

 extinct genera than Phenacolestes arises again. 



Whether Megapodagrion and its near ally Allopodagrion are to 

 be looked on as descendants of Phenacolestes or of Melanagrion, 

 18 



