234 PROCEEDINGS OF THE ACADEMY OF [May, 



neura (1897), referred it to the Calopteryginffi. In 1902, I placed 

 it in the legion Podagrion of the Agrioninse.^'' In 1903, Professor 

 Needham associated it with the Anormostigmatini ( = legion 

 Pseudostigma of Selys), stating that "it has all the essential vena- 

 tional characters of this group, and surely these are sufficiently 

 distinctive."^^ A comparison of Thaumatoneura with these essential 

 venational characters'^ gives the following result. 



Thaumatoneura agrees with the Anormostigmatini in that Us 

 separates from M1+2 at or very close to the subnodus and the part 

 of the wing beyond the nodus is greatly enlarged. On the other hand, 

 Thaumatoneura differs from the Anormostigmatini in that, while the 

 stigma is not braced, it is not diffuse or lost ; the matching of cross- 

 veins is not so marked; the longest vein between Mi and M2, which, 

 in the preceding pages, we have assumed to be Mia, does not parallel 

 M2 more than it parallels Mi and lies almost midway between Mi 

 and M3; it has shorter supplementary sectors between itself and Mi 

 and also between itself and M2; these latter tend to be longer and 

 better developed than those anterior to Mia, and in this Thaumato- 

 neura resembles Paraphlehia and differs from the Anormostigmatini. 

 This tendency to the development of more rows of cells between 

 Mio and M2 than between Mi and Mia is emphasized in Philogenia 

 carrillica, P. terraha, Heteragrion tricellulare, H. erythrogastrum and 

 Argiolestes icterotnelas which have no supplementary sectors between 

 Ml and Mia, but two between Mio and M2; all these are Podagrionines. 

 Even in more reduced venations of the same group, as in Allopoda- 

 grion there are no supplementary sectors between Mi and Mia, but 

 still one between Mia and M2. On the other hand, even in Anormos- 

 tigmatini with the most reduced venation, e.g., Mecistogaster, there 

 is a longer supplementary sector between Mi and Mia than between 

 Mia and M2. Professor Needham himself has pointed out some 

 further differences of Thaumatoneura from Megaloprepus, and still 

 others may be added. Thus the nodus and quadrilateral (quad- 

 rangle) are no more approximated in Thaumatoneura than in many 

 other undoubted members of the legion Podagrion (See Table I, 

 p. 250, posted). Thaumatoneura possesses more supplementary 

 sectors between M2 and Rs and between Rs and Ms. I believe, 

 therefore, that Thaumatoneura shows more resemblances to the 

 legion Podagrion than to the legion Pseudostigma, but it does not 



i« 1902, pp. 29-32. Biol. Centr. Amer. Neur., p. 58. 

 " Genealogic Study, pp. 748-749, footnote. 

 18 Needham, I.e., p. 748. 



