214 
[October 
stant.* All this is generally expressed by saying that so and so are 
difficult, variable or polymorphic groups; but I do not know that any 
one has yet called attention to the fact as a confirmatory proof of the 
Derivative Origin of species. Mr. Darwin certainly has not done so. 
The question seems unanswerable :—If species were separately created, 
not derived one from the other, why should variability have been 
largely conferred upon some entire groups and almost entirely denied 
to other entire groups ? Why should the spotted Erotylidge be con¬ 
stant, and the spotted Coccinelladae inconstant, in their spottings? 
Throughout Odonata, as is well known, the S abdominal appendages 
are used to embrace the neck of the 9 preparatory to and during copu¬ 
lation, and for that purpose they are in general curiously curved and 
in many cases, especially the two superiors, armed with teeth thorns or 
branches. The 9 abdominal appendages are here, on the contrary, 
always simple—laminate, elongate-conical or cylindrical—and almost 
invariably smaller than those of % . In most genera of Locustarige 
Latr. (=Grryllid9e Leach,) e. g. in Phylloptera, Orchelimum, Xiphi- 
dium and Conocephalus, but not in Rhaphidophora*}* which osculates 
In my Paper on III. Pseudoneur. (p, 381) I stated erroneously that the milky- 
blue tint {hleu soupoudre) often found on the bodies of certain mature Libel- 
lulina ‘-'seems to be secreted under the external integument, and what is 
known as pruinoseness in Agrionina on its surface, whence it may be washed 
off.” In both tribes the bluish matter is secreted externally, and in both it 
may be easily washed off. Hence variability arising from this source can 
scarcely be considered as eolorational variability, and so far as regards this 
character the tribe Lihellulina are not variable in their coloration. 
f The tetramerous Locustarian genus Ehaphidophora osculates through the 
partly trimerous allied genus Haihinia, Hald., and the partly tetramerous Gryl- 
lide genus CEcanthus, with the elsewhere uniformly trimerous Gryllides. Hence 
it is not surprising that the'^ 9 abdominal appendages should here be antenni- 
form as in Gryllides. Ehaphidophora and its allies agree also with Gryllides, 
except Tridactylus, and differ from all other Locustarise known to me, in the"^ 
not having any inferior abdominal appendages. In Loeustarise, with the above 
exception, in Blattadse, in Mantidee, and in Acridii, the % but not 9 has two in¬ 
ferior appendages. In Blattadae and Mantidse these appendages are separated 
from each other, as in the tribe Agrionina, and antenniform, and are known as 
‘‘anal styles.” In the Locustarian genera Phylloptera, Orchelimum, Xiphi- 
dium and Conocephalus they are laminiform and soldered together more or less 
at base and antenniform at tip. Finally, in the Locustarian genera Cyrtophyl- 
lus and Phaneroptera and in all Acridii, they are soldered together either 
