2 Cimicidae. 
included in the table, as their wing-structure is un- 
fi ! li e not Se 
AUS alerseities “Discocephalinae” and “Graphosominae” are treated 
= ? 
known. The “Acanthosominae”, 
as tribes of the Pentatominae. 
The accompanying tree expresses to some extent my idea of the phylogeny of 
the Cimicidae. 
10 Tessaratominae 
9 Coptosominae t 
etude 
7 Seutellerinae ee 
Sy GE zy 
~ ° Discocephalini Acanthosomini 
6 Cyrtocorinae 8 Aphylinae + 
Graphosomini | | 4 Phioeinae 3 Phyllocephalinae 
tT tT 
i 
| 
2 Pentatominae 
1 Cimicinae 
The Cyrtocorinae may have sprung from a Graphosomine stem before the 
branching off of the Scutellerinae, or from a Scutellerine stock. The Tessaratominae 
may have sprung direct from a Pentatomine stem. 
My reason for treating the Acanthosomini and Graphosomini as merely tribes 
of the Pentatominae is, that I can find no valid characters upon which to raise them 
to the rank of subfamilies. As regards the Graphosomini the development of the 
scutellum is of no importance, as otherwise, Stiretrus and Discocera should form a 
subfamily; the other characters mentioned by authors are by no means exclusive. ~ 
As regards the Acanthosomini, Stal has given no exclusive characters, as certain 
true Pentatominae have bisegmentate tarsi. Duda (1885 Wien. E. Z. IV. 99) places 
them as a subfamily “because of their anatomical peculiarities’, but he seems to 
have based his arguments on one or two species only, and I do not think his re- 
sults are valid. 
