2 JOURNAL OF ENTOMOLOGY AND ZOOLOGY 
there have been two well-marked lines in the phylogeny of the 
suborder, the Tenthredinid and the Siricid stems. The Xyelidz 
and Pamphiliide are very primitive forms that represent off- 
shoots from near where these two stems divide. The Siricide 
also retain many highly primitive characters, although in other 
respects ‘‘sidewise specialized’’. The Xiphydriide, Cephide, 
and Megalodontide group themselves with them. The Oryssidx 
represent the most highly modified group within the suborder. 
They are more divergent from any other family than are any 
of the other families from each other. Yet I believe they had 
an ancestry somewhere along the Siricid stem. I am not con- 
vineed of the taxonomic advisability of erecting super-families 
for small groups of their nature, representing as they do, 
highly specialized offshoots of some other stock. 
The classification offered by Dr. MacGillivray (1906) was 
based upon careful and critical comparative study of a single 
set of organs—the wings, and seems more conservative and 
more in accordance with my own views. I have followed, in the 
main, the arrangement which he proposes. 
I am not prepared, from personal knowledge, to offer an 
opinion upon the advisability of dividing the Tenthredinida 
into several families, as is done by Ashmead and Rohwer. It 
is outside of the scope of this paper, and I have followed Dr. 
MacGillivray’s classification in this regard. 
THE SUBORDER CHALASTOGASTRA 
A Key to tHe F'AmMIies 
A. Front wings with R, present, possessing three marginal cells. 
XYVELIDH 
AA. Front wings with R, absent, therefore possessing one or two but 
never three marginal cells. 
B. Front wings with subcosta present as a distinct longitudinal 
vein. PAMPHILIDA 
BB. Front wings with subcosta absent. (Rarely it is present as a 
pale, very indistinct line, closely appressed to R-+ M, or 
Se, may be present as a transverse vein). 
