PaiUmd.] 354 [May 2, 
ancestors of tlie three orders just mentioned, while we have 
endeavored to show tliat the Coleoptera may liave diverged from 
sonie metabolous neuroj^terous form.^ 
It is not improbable that the original form of the larval Diptera 
was more or less eruciform, with a large head, roomy enough to 
contain the brain and infraoesophageal ganglion, with a definite 
clypeus and labrum, as well as a mental region, and with mouth- 
parts much more generaUzed and equally developed than in any 
existing form. The adult must have had a less highly con- 
centrated thorax than in any existing form, the metathorax 
being more completely developed, and bearing a pair of wings'^ ; 
perhaps, as in the Hymenoptera, smaller than the anterior pair. 
The abdomen was probably not less elongated than in the existing 
Tipulidae, this family, with the Bibionldae, Cecidomyidae, 
Mycetophilidae, etc., perhaj^s being the most primitive and 
generalized of existing Diptera. 
The larva of the Siphonaptera apparently presents the nearest 
approach of any of the insects now existing to the shape of the 
primitive Diptera. It is certainly a more perfectly developed 
larva as regards its external structure and also the position of the 
brain than anj^ dipterous one, and approaches nearer to our 
conception of the j^rimitive, ancestral, dipterous larva than anj'^ 
other form. 
Thus far no traces of temporary abdominal legs have been 
observed in the embryos of either the Siphonaptera or Diptera, 
and «as I have before remarked : "The lack of these structures in 
dipterous embryos appears to confirm the view that they are the 
most extremely modified of all insects."^ 
As regards the phylogeny of the Siphonaptera we can only 
say, with our present imperfect knowledge of their embryology, 
that tliey seem to stand nearer to the Diptera than to any other 
order, and that they must have diverged from the ancestral 
1 Third report U. S. entomological commission, p. 299, 1883. Also Amer. iiat., 1883. 
- Kudimi^nts of tho second pair of winj;s are said to exist in the pupae of some of 
the more generalized Diptera, and the general resemblance between the pupa of 
Diptera and of the Tineina suggests a* certain relationship between the moths and the 
tlies. 
^ A study of the transfonuMtions and anaton-.y of /.iii/oa rri.y/dld. Vvuv. Amer. 
philos. soc., V. ;j-J, ]). '28(i, 1894. 
