87 
a gap at the inner side, a condition almost identical with that found in 
Sesia ; but in many of the species (or chiefly among the Pamphilidi ?) 
there is a complete circle, repeated in three rows, the outer row having 
the smallest hooks; this latter form is met with elsewhere only in 
Hepialus, and is, therefore, a very archaic form of pro-leg. 
In the pupa there is one “ Micro ” character to which I find that 
Scudder calls attention, though he does not appear to be aware of its 
significance—the persistence of the dorsal head-plate (cephalo-thoracic 
piece). In addition to this feature, the pupa presents several others 
that point to its still strong affinities with the Micro-Heterocera. I 
have only possessed a few Hesperid pupae, and those which I have 
examined from other collections had of course to be respected, so that 
I am still in some doubt upon some of the points connected therewith. 
I think it is the case, that the species belonging to Scudder’s tribe 
Pamphilidi, which includes our sylvanus, linea, &c., possess in a more 
marked degree the features of the Micros than do those belonging to 
his Hesperidi. I will, however, only say that some, and I think pro¬ 
bably a majority of, species present most of the “ Micro ” characters 
which I am about to enumerate. I need merely refer to such well- 
known characters as the making of a cocoon, and hybernation as a full 
or nearly full-grown larva ; both of these are distinctly moth characters, 
and the latter is a “ Micro ” habit and does not, I think, occur in any 
true butterfly. The possession by the pupa of a dorsal head-plate, as 
noted by Scudder, is also a “ Micro ” character, and is associated with a 
still more notable one which appears to have escaped his attention— 
the separation of the eye-plates (which are dorsal), on dehiscence, from 
the ventral head parts, and their continued attachment to the dorsal 
head-plate.' Another most remarkable “ Micro ” character, for which . 
I was not at all prepared, is the persistence in the pupee of some species 
of the terminal joints of the maxillary palpus as a minute “ eye-collar.” 
This, and the following, however, do not occur in any species of which 
I have had specimens in my own possession, and I therefore mention 
them with reserve. 
Another character is the persistence of the 7th abdominal segment 
as a free one in the male pupa: this appears to be the case in some 
species of Pamphilidi, but I do not like to be too positive, as one cannot 
be quite sure on the point with an empty case, unless one is at liberty 
to dissect and mount it. It is at least certain that the incision appears 
so open after dehiscence, that if not actually movable, it can only 
recently have lost its mobility. 
These Micro-characters, together with the larval prolegs pointing to 
Hepialus as a possible ancestor, lead one to ask whether there may be 
any intermediate forms that mark the steps by the way. Hepialus is 
now often instanced as a very archaic form, even as the possible 
ancestor of eveiything; Hepialus is, however, a terminal form ; that is, 
it is at the end of its line of evolution and has no descendants. This is 
tolerably clear from its minute antennas, but especially from the 
circumstance, remarkable in so early a form (indeed it is the only 
instance amongst unquestionable “ Micros ”), that the male pupa has 
the 7th segment fixed. 
The real ancestor that is intended, when Hepialus is referred to, is 
the common ancestor of Hepialus, Cossus, &c. Hepialus retains sundry 
archaic features of this common ancestor more persistently than Cossus, 
