1890 .] 
87 
[Packard. 
It may here be observed that the abdominal or membranous legs 
of Lepidoptera are a very ancient heirloom and may have been 
handed down by the Scolopendrella-like ancestor of the insect- 
phylum. While there are rudimentary membranous legs in certain 
dipterous larvae as Ephydra, they are far more perfectly devel- 
oped in Lepidoptera than in other insects. The slight tendency 
shown towards segmentation in these legs may either be due to 
reversion or to muscular strains in a comparatively recent structure. 
In either event, the incipient folds of the integument suggest 
the probable mode of origin of the normal, or thoracic legs. The 
latter may have been, at first, like the rudimentary membranous 
legs of Peripatus and of certain dipterous larvae; another step 
would be the membranous, but bookless abdominal legs of the lar- 
val Tenthredinidae ; and a third advance, the abdominal legs of cat- 
erpillars. If the integument of the primitive limb should be farther 
hardened by an increased secretion of clritin, the mechanical ne- 
cessity of freely moving membranous joints would result, and this, 
accompanied b} r a reduction of the hooks tp a single pair, would re- 
sult in the normal insectean limb. 
As to the primitive number of crochets, with the limited amount 
of alcoholic material at my command, I have thus far found four 
onty in the freshly hatched larvae of Orgyia and Parorgyia, and 
as has been observed, they are arranged in two pairs or sets, with 
a wide interval between the sets, which is nearly filled up by the 
second growth of hooks at the first moult. Fig. 5 represents the 
planta of the larva of Parorgyia parallela after the first moult. It 
will be seen that the number of hooks of the first stage has been 
trebled, there now being twelve. Yet they are still arranged in 
two sets, with a slight space between the sets, and this arrangement 
appears to persist throughout all the larval life of the insect. 
Though the grypogene is not represented, I think I have detected it 
in some mature Bombycine larvae. An interesting feature is the 
single series of minute bulbous tubules fringing the edge of the 
planta (fig. 5 a) and which recall in general appearance the gland- 
ular or tenant hairs of the pulvillus of the feet of flies and other 
insects. 
In the Lasiocampidae, which are generally placed at the foot of 
the group of Bombj^ces, the number of hooks is eight. In the 
freshly hatched Clisiocampa americana (fig. 6) there are three di- 
visions of the leg, and the eight claws are arranged in a curvilinear 
