69 
transverse strip divided in the middle, and the rein is an oval projection 
beyond this. Within it is visible in some specimens a tubular structure, 
very similar to that seen in such forms as Callophrys rubi, where it is 
quite internal, not in an external structure as here. (Plate II., fig. 4.) 
Lycaenopsidi ( argiolus ) have a very similar structure. (Plate II., 
fig. 2.) 
In Lampididi ( Baetica) these parts are much larger, so that their 
extremity, when extended, reaches beyond the end of the ovipositor, 
whereas, in the three tribes previously noticed, it does not extend 
beyond the end of the 9th segment, the terminal portion (rein) is 
rather elaborate in structure, and the whole organ gives the impression 
that it can be invaginated in a very similar way to that which occurs 
in the Plebeidi. I do not think this can occur in Everidi, but in the 
other tribes it may perhaps do so in some degree, but what occurs in 
them is rather perhaps retraction than invagination. (Plate II., fig. 3.) 
In the Plebeidi we have this structure carried to an extreme. The 
basal portion (the prop) curtains a chitinous loop which, when extended, 
reaches to about the end of the ovipositor, whilst the terminal portion 
(the rein) is even longer, is of unchitinised tissue, except a plate at its 
extremity. When at rest and invaginated, the rein passes inwards, 
without change or inversion, and to receive it the prop turns outside 
in, so that in this position the terminal chitinous plate is level with 
the opening behind the 7th abdominal segment. (Fig. 5.) 
I hope to obtain some evidence as to the precise functions of these 
several parts, about which at present I have only some doubtful 
guesses. 
Reverting to the female structures of the Plebeiids, as more easily 
examined, we find the usual double terminal portion with its tactile 
hairs and its two rods, the outer rods. Behind this, dorsally, is a 
plate, with plain indication of the inner rods. It is not usually clear, 
what, if anything, corresponds to this ventrally, but preparations of 
thetis and amanda, in which certain ventral scaling is left, suggest that 
the segment is completed ventrally in membrane without any chitinous 
plate. Behind this we come to the 7th abdominal segment, which is 
unmistakable, both by counting from the base and by its possessing 
spiracles. 
When the parts are at rest and undisturbed, this is all we can 
observe—segment 7 and dorsal portions of only two segments beyond. 
But ventrally, just beyond or within segments 7, are parts that 
unquestionably represent portions of the 8th segment, but, practically, 
there is nothing to be seen of the 8th as a definite segment, or of the 
ventral plates of the 9th and 10th segments. 
What has become of the external surface of the 8th abdominal 
segment ? 
This brings us to a very difficult subject, which I don’t profess to 
have solved, or even very clearly to understand, viz., the segmentation 
of the last portion of the ? abdomen of Lepidoptera. 
The abdomen of Lepidoptera consists of ten segments. In the male 
imago, these are easily verified, there being eight obvious segments, 
and the genital appendages absorb the two others, which can often be 
distinguished from each other without difficulty. In the female there 
xxii.-xxiii. 
