162 
membrane, which, however, does not reach the base of the triangle. The 
remaining chitinous area is sharply limited in front of this membrane, 
apparently by a vein, which is perhaps media much displaced, although in 
some cases the remnants of media seem still present on the anterior side 
of the claval furrow, the narrow clavus remaining chitinized for a large 
part of its length, bounding the membranous area posteriorly. The 
Graphosomatinae, also with an expanded scutellum, have the corium simi- 
larly reduced, but the claval region is also membranous except at base, 
the chitinous portions being delimited by a very oblique, nearly straight, 
line. 
The hind-wing venation is more generalized and of considerable 
systematic importance. The folds are easily confused with the veins, 
but are recognized as being most distinct at the outer margin, where they 
end at the retreating angles between the more or less scalloped lobes. The 
first furrow, that behind media, corresponds to the posterior one of the 
fore wing. 
By far the most primitive hind wing I have seen in the Pentatomidae 
is that of one of the Coptosomidae (Brachyplatys? sp.) from Manila. 
The stems of subcosta, radius, and media are all complete and apparently 
simple. In the next lobe, between folds, the two short veins—which in 
Pentatomidae are often connected basally near the middle of the wing but 
have no common stem on the basal half of the wing, or at best only a 
doubtful one—seem to be the posterior branch of media and the anterior 
branch of cubitus respectively, cut off by the folding of the wing. Behind 
this are two more very distinct lobes: the first contains cubitus and the 
first anal; the next and last, the second anal. The strongly lobed wing- 
outline closely resembles that of Thyreocorinae, but there is no trace on 
cubitus of the stridulatory organ found in all Cydnidae examined (PI. 
XVI, Fig. 6, 8). Radius and media are fairly straight, and are con- 
nected by a long discal cross-vein. 
In the Scutellerinae (Homaemus proteus, Pl. XVI, Fig 5) the fol- 
lowing modification appears, which I have found in all related Heterop- 
tera: radius sags down at the cross-vein to meet media, and media has an 
upward sinus, thus shortening, nearly obliterating, the cross-vein. In 
other forms the two stems are usually fused for a short distance. The 
basal part of radius as far as the beginning of the sinus, closely parallel 
to subcosta, is very faintly or not at all traceable; but the decurved part 
persists as a strong curved spur, the hamus, free at its basal end. The 
second vein in the middle fold is apparently wanting. The wing is evi- 
dently related to that of the Coptosomidae. 
In Coreidae (Anasa tristis, Pl. XVI, Fig. 10) radius is feebly indi- 
cated basally, and discally is fused for a short distance with media. An 
evident hamus is present. Otherwise the venation is essentially as in 
Pentatomidae. 
In the Thyreocorinae (Pl. XVI, Fig 11), in the Halydinae, in Tes- 
saratoma (Fig. 7), and probably also in Cyrtocorinae, radius seemingly 
is present, slightly diverging from subcosta. But the clear-cut hamus 
