ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 
589 
its own, and that traces of an extremely remote past can be deciphered 
by the study of its structure. 
In investigating the persistent traces of larval structures upon the 
pupa, the author describes the claspers, the caudal horn of the Sphin- 
gidse, and other structures, the larval tufts of hair indicated on the pupa, 
and the larval markings. He next discusses the number of abdominal 
segments and their relation to those of the larva, and concludes that 
both possess ten abdominal segments ; even if this be shown to be 
incorrect it will not affect the segmental relations of the external 
reproductive organs, for they only come into relation with the eighth, 
ninth, and ventral (anal) part of the tenth abdominal segments. 
The external generative organs are next described, and it is sug- 
gested that the median prolongation of the tenth abdominal and the 
relation of its apex to one of the generative apertures represents an 
ancestral ovipositor, now represented only by its external cuticular layer. 
The relation of the pupal to the imaginal antennae, and the history 
of the degeneration of the antennae in female imagines form the subject 
of the next (fourth) part of the memoir; and the pupal wings are after- 
wards examined. In the course of an interesting discussion it is pointed 
out that when the two sexes seem to approach most closely in competition, 
flying together and both apparently exercising the powers of active 
selection — when, in fact, courtship appears to be mutual — then the 
differences between the antennae of the two sexes become very small, 
and in the cases of most complete equality disappear altogether. The 
antennae are in all probability sense-organs of very general use, although 
their sexual function is by far the most important, while free and active 
flight gives abundant opportunity for their exercise in all possible 
directions, so that these organs may be sometimes equally developed in 
the two sexes. 
Phylogeny of Lepidopterous Larvae.* — Prof. A. S. Packard 
publishes some very interesting observations on the larvae of Lepido- 
ptera ; his studies have led him to believing, provisionally at any rate, 
that the butterflies have originated from moths w r hich resembled the 
Bombyces more than any other group ; at any rate the ancestors 
were hairy or spiny caterpillars. The Nymphaliidae may have originated 
from Arctian-like forms, and the Papilionidae from Attacids. They 
certainly show no signs of descent from the Sphingidae, the Castniidae, 
Agaristidae, Cossidae, or Hepialidae. 
The Hesperidae appear to be the most generalized butterflies, but 
their origin is not apparent. The Papilionidae probably stand next 
above them, as they seem to have descended from an earlier and lower 
type than the Nymphaliidae. The Lycaenidae appear to be the most 
extremely modified, and form a shoot perhaps parallel to the last; they 
are a more modern and highly modified family, though somewhat 
degenerate as regards their larval form; they thus recall the 
Cochliopodidae, which are highly modified Bombyces. 
Sound-Organs of Dytiscidse.t— Herr P. Becker, after a historical 
account of our knowledge of the sound-organ of Pelobius, a genus of 
* Proc. Boston Soc. Nat. Hist., xxv. (1891) pp. 82-114 (2 pis.), 
f Arch. f. Naturg., lvii. (1891) pp. 105-12 (1 pi.). 
