72 
after we leave Eriocephala they belong, as a rule, to a dorsal and not 
to a ventral plate. This, however, is not an important point, as not a 
few species have the rods bifurcating, one branch going to the dorsal, 
one to the ventral plate. A not infrequent form of this arrangement 
shows the two branches, one passing to the forward margin of the 
dorsal plate, the other to the posterior angle of the ventral plate carry¬ 
ing the pairing opening, clearly showing the dorsal plate to belong to 
the same segment as the ventral portion to which it is attached, but 
does the ventral portion belong to the 8th segment, or is it the 9th 
segment attached to the 8th? If it belonged to the 8th one would 
expect it to be attached to the forward margin, not close to the posterior 
margin. This attachment to the posterior margin of the 8th (-p9th ?) 
segment iswell seen in the photograph of Tinea pdlionella (pi. III.,fig. 1). 
The pairing opening, however, so modifies the 8th segment that it is 
difficult, and, indeed, often impossible to determine its parts. 
I have not any photographs of this arrangement to show you, 
except one of Erynnis ello, a common South American sphinx. With 
this I show you a photograph of the 8th ventral of the male imago, 
which has interested me a good deal, as appearing to be a transmission 
to a male of a female sexual character, it looks as if the forward 
margin of the segment was influenced by a tendency to form a canal 
like that of the 5 . 
The tendency of secondary sexual charactei's to be transferred to 
the other sex, is not, I think, uncommon, and I believe I have observed 
other instances, and Darwin recognises it as a not very common, but 
still ordinary, circumstance. The chief interest is this case is that the 
feature transmitted is rather a primary than a secondary sexual 
character. This segment is similarly affected in other Sphinges, but 
not so markedly in the few I have examined as in ello. 
A species in which the inner rods divide and go to the inner 
margin of a dorsal plate and to the outer margin of apparently the 8th 
ventral segment is Euxanthis zoeyana, in which this structure is very 
obvious. 
All this rather points to the inner rods belonging to the 8th segment 
and that the 9th segment is the missing one. I submit, however, an 
interpretation that takes account of the position of the ventral attach¬ 
ment of the inner rods. 
When, however, we explore in some other directions, this is less 
obvious. Some Tineas and Pyrales give the most direct indications 
that the missing segment of the 8th, excepting so much as carries the 
pairing opening, and that the 9th, is well developed and possesses the 
inner rods. I do not possess any photograph taken to demonstrate this, 
but show you some taken for other purposes, one of Tinea pellionella 
and a series of Scoparias. (PI. III.) One of the most convincing 
preparations I have is of OEcophora pseudo-spretella, which shows the 
pairing pocket withdrawn into the 7th abdominal segment, the rest of 
the 8th segment being missing, whilst beyond this the inner rods divide 
and the branches join the inner margins of the dorsal and ventral 
plates of what can be only the 9th segment. The outer rods going as 
usual to the terminal (10th) segment. (PI. V., fig- 1-) 
A cursory glance at this would make one believe that the pairing 
xxii.-xxiii. 
