388 FRANCIS DARWIN. 
atalanta. The chitinous parts forming the channel and the 
rest of the proboscis are very much stronger in Ophideres 
than in Catocala, and the muscles which serve to extend 
the proboscis and make it into a stiff rod are more developed. 
The proboscis of Ophideres is apparently unable to roll up 
into the close helix which we find in Catocala and most Lepi- 
doptera. Each half of a Lepidopterous proboscis is said (New- 
port, loc. cit.) to be made up of a number of superimposed 
rings, and in the proboscis of Catocala five transverse lines 
are seen marking out these rings. In Ophideres, however, 
these file-like markings are oblique instead of directly trans- 
verse, and at a point (p, Fig. 2) they are abruptly bent at an 
acute angle, making a beautiful series of V’s on each side 
of the line (g); the point (p) is shown also in the section 
(Fig. 3). The oblique rings give a curious imbricated appear- 
ance to the periphery of each half of the proboscis. 
WiGaes 
Fic. 3.—Transverse section of the proboscis (not the sawing part. Semi- 
diagrammatic). 
ch. The channel or tube of the proboscis. 
tr tr. The large trachea running down each half of the organ. 
hh. Locking apparatus, keeping the two halves in contact. 
g- Groove running down ventral surface of proboscis. 
dsp. Shows the position of the dorsal frill of spines; these are 
so delicate that nearly all are broken in making sections. 
m. The space filled up by muscle which is here omitted. 
p. Corresponds to p in Fig. 2. See text. 
Finally, the apparatus which locks the two halves of the 
organ together is very remarkable. Fig. 3 is a section of 
the proboscis showing the strong curved teeth (A, h) which, 
by closely hooking into one another, must effectually pre- 
vent dislocation; the delicate spines (dsp) on the dorsal 
surface may perhaps contribute to the same result. Neither 
