XXI 
Classification. 
I have recently discussed the phylogeny of the Heteroptera in another 
place’), but as the classification adopted here, differs rather widely from 
those employed by previous authors, a little explanation is necessary. 
Reuter *) says: “J’ai commencé mon ouvrage [Hemiptera Gymnocerata 
Europae] par la famille des Capsides, parceque celle-ci me parait étre la plus 
basse dans le systeme.” On the contrary, the “Capsidae” appear to me to 
be very highly specialized, as nearly, analogically of course, in the direction 
of Hymenoptera and Diptera as their Hemipterous constitution will permit %). 
To me, it appears, almost self-evident that the Pagiopoda are derived 
from the Trochalopoda, or are at least, less primitively Heteropterous. The 
Notonectoideae also are derived from the Miroideae. The Nepoideae are 
almost certainly derived from a Myodochoid stem — or one very close 
thereto -- and the only serious difficulty is in the relative positions of the 
Myodochidae (or rather Pyrrhocoridae, for the Myodochidae seem to have 
been derived from ocellate Pyrrhocoridae) and Cimicidae. 
The Schiedtean division of the Heteroptera is practically accepted 
by all of the formost workers *). Each of his two divisions seem to be further 
divisible into two superfamilies °); while these are pretty self-evident, it is 
difficult at present to limit them sharply. They may be grouped as follows: 
Phalanx Superfamily Families 
| Cimicidae; Thyreocoridae; Urolabi- 
1. Cimicoideae didae; Aradidae; Coreidae; Pyrrho- 
coridae ; Myodochidae ; Tingidae. 
Nabidae; Naeageidae; Hydrometridae; 
Gerridae; Reduviidae; Macrocepha- 
lidae ; Enicocephalidae ; Nepidae. 
1. Trochalopoda | 
2. Nepoideae 
Anthocoridae ; Clinocoridae ; Polycte- 
3. Miroideae nidae; Aépophilidae; Miridae; Dipso- 
| coridae. 
2. Pagiopoda ' Acanthiidae; Ochteridae; Nauco- 
4. Notonectoideae ¢ ridae; Belostomidae ; Corixidae ; 
Notonectidae. 
The following tree is slightly different from that recently published, 
but the points of variance are likely to remain doubtful till more is known 
of the earlier stages. Nothing is more difficult to construct than a phylo- 
genetic tree, but I think that the main lines of this are approximately 
correct. 
1) 1908 Canad. Entom. XL. 357. 
2) Act. Soc. Sci. Fenn. XIII. 6. ; , : 
8) Reuter now agrees that the “Capsidae” are a terminal, highly specia- 
lized family. é 
4) Reuter, 1908 Mém. 8. E. Belg. XV. 90, does not now accept it. 
5) Kirkaldy 1907 P. Haw. E. S. I. 137—88. 
