1947] 
Banks — Characters in the Perlidce 
275 
cubitus; I have chosen the latter course. Bristles are 
ancestral, and the occurrence of ancestral characters in 
descendants is not unusual. 
I had started this paper in order to do identification, 
especially of New England Perlidae, and had hoped to rely 
chiefly on body characters, but find (as others have) that 
there is much in the wing, particularly the basal portion 
that is sufficiently constant for synoptic purposes. I do 
not underrate male genitalia and am pleased to find that 
in general my results agree with the plan made by Ricker 
in 1943. 
Table to the Subfamilies of the Peblid^: 
1. Head longer than broad, eyes situate more than twice 
their diameter from hind margin; in the forking 
of cubitus the upper branch is bent a little from the 
stem; the third anal vein arises from the cell; no 
anal lobe nor anal brace ; antennae with a few joints 
near base short; a carina or line connecting eyes 
and passing just behind hind ocelli. 
Kathroperlinae 
If head is longer than broad the eyes are not more 
than one and one-half diameter from the hind 
margin 2 
2. Small forms less than an inch; no anal lobe nor 
brace ; anal area of hind wings with not more than 
four longitudinal veins ; distinct erect bristles in 
a row on each side of femoral groove ; occipital 
line absent or indistinct ; antennae rather long, not 
thickened at base, the joints moniliform; usually 
in the fore wings the third anal vein branches from 
the second beyond the anal cell Chloroperlinae 
Anal area of hind wings with more than four longi- 
tudinal veins ; in the fore wings the third anal vein 
leaves the second at or before end of anal cell 3 
3. But two ocelli 4 
With three ocelli 6 
4. Head about twice as broad as long, partly under the 
overhanging pronotum, latter very broad, rounded 
behind, straight or slightly concave in front; legs 
