LEPIDOPTERA. 195 
veral have no proboscis. Some of the females are destitute of wings, 
or have but very small ones. The caterpillars most commonly spin 
a cocoon; the number of their feet varies from ten to sixteen(1). 
The chrysalides are always rounded, or without angular elevations 
or points. 
The classification of this family is very embarrassing, and with 
respect to it our systems are as yet merely imperfect essays or rude 
sketches(2). We divide it into ten sections. The first nine are 
composed of those species in which the wings are perfectly entire, 
or without digitations. All those that in their caterpillar state 
live almost exposed or in fixed domicils, several of which have at 
least sixteen feet, and which, in their perfect state, have their supe- 
rior palpi very small or entirely concealed, the wings more or less 
triangular, horizontal or tectiform and not folding round the body, 
will compose the first eight. The last of these latter, or the eighth, 
is the only one in which the caterpillars have fourteen feet, two of 
them anal. If the same number be found in some others, then the 
two last are wanting. 
The two divisions ftacus and Bombyx of the genus Phalena of 
Linnzus correspond to the four first sections. The proboscis is 
most frequently rudimental or very small, and its two threads are 
not united. The inferior palpi, those of a small number excepted, 
are small and almost cylindrical. The antennz, at least in the 
males, are pectinated or serrated. The wings are horizontal or 
tectiform, and in several the inferior ones project beyond the others 
when at rest, and sometimes are also destitute of that bristle or 
bridle which connects them with the latter. The thorax is always 
smooth, as well as the abdomen, and woolly. ‘The latter is usually 
very voluminous in the females. The cocoon of the chrysalis is 
usually well felted and solid. 
Although the Nocturna of the fourth section are closely allied to 
those of the preceding ones, we find a character in their caterpillars 
(1) De Geer, in one species, counted eighteen, all membranous, II, p. 245, and 
I, xxx, 20; xxxi, kS—16. 
(2) Weare frequently compelled to borrow characters taken from the cater- 
pillar. If this be disregarded, we shall be compelled to suppress a great number 
of genera. I will mention for instance that of Phalzna proper, or Geometra. 
If we consider only the perfect Insect, it is impossible to distinguish generically 
several species, such as the prodromaria, betularia and hirtaria from Bombyx; it is 
also evident that we could not separate from them Platypterix and other genera. 
