WING-VENATION OF THE PLECTOPTEKA. 161 
notation, for the simple reason that its authors themselves admitted that it 
■vvas unsatisfactory ; and, later on, they accepted Miss Morgan's interpretation 
instead (see Comstock, 1918, chap. x.). We have, then, only to consider 
Miss Morgan's notation. With respect to this, we may note the following- 
points:— - 
(1) The. research was carried out on wings oE the last larval instar only. 
In this instar the tracheation is the most variable of all, and the posterior 
continuation of the alar trunk is usually absent. In spite of this, Miss 
Morgan correctly interpreted the homologies of the cubital trachese, but 
failed, for some reason, to connect them with their true venational homologues 
in the imago. Had she examined the tracheation in this region in earlier 
iustars, she could not have failed to note the close correspondence between 
the strongly bent cubitus in the larva and imago ; and her results, as far as 
the limits of Cu were concerned, would then have agreed with those given in 
ihis paper. 
(2) The evidence for her chief innovation; viz. the crossing of a simple Rs 
over two branches of M, as in Anisopterous Dragonflies, is not forthcoming 
from the great majority of the larval wings examined, but only from the 
wings of a single species, in which, as she admits (I. c. p. 98), " half of 
ilie wing-pads showed the radial branching just described and half of them 
.gave no sign of it." The species which gave, in only one-half of its 
representatives, an indication of Rs crossing over M, was a species of 
Heptagenia, a genus which is by no means the most archaic of the fifteen 
genera studied by Miss Morgan, Yet she accepts this evidence, against the 
weight of evidence supplied by all the other fifteen genera and the other 
species of Hejttagenia included ! 
As I have already pointed out, new light on the venation of Dragonflies 
has made it certain that Rs does not cross over M in the Zygoptera, and has 
also rendered it extremely improbable that the trachea which does this in the 
Anisoptera is Rs at all. The evidence for a crossing of Rs over M in 
Mayflies to be found within the Order itself is, as we have seen, practically 
negligible. Thus Miss Morgan's whole case totters to the ground, and must 
be replaced by a more lasting structure. 
To sum up, we have seen how the evidence from the Convexity and 
<Joncavity of Veins agrees absolutely with that offered by a study of the 
Permian fossil Protereisma, and the combined evidence of these two lines of 
■study is supported fo a very great extent by that of the tracheation of larval 
wings, the only doubtful point in this hitter evidence being its undoubted 
tendency to vary. Taken all together, the new system of, homologies appears 
to me to be supported by so great a weight of evidence as to be at once 
.■acc'eptable in place of the hypotheses previously offered in the Comstock- 
Needham and Morgan notations respectively. At the same time it offers a 
-new and clear basis on which to build up a system of classification within the 
11* 
