180 [June, 1891. 



forwards on the sheath or 8th segment ; in other cases, again, it is 

 developed into a huge plate reaching backwards under the 7th segment 

 as far as its inner edge, and forwards to the outer edge of the 8th, 

 with the underpart of which it is intimately united, the orifice being 

 occasionally guarded by a flap, chitinous or leather}^ as the case may 

 be ; whilst the varieties and modifications of the several forms are 

 infinite, and are likely to afford useful characters in determining 

 closely allied species. In other genera of the Cramhidce, as well as in 

 all the remaining great divisions, with the exception of the Sphingidce, 

 which I have had no opportunity of examining, the lower rods, so far 

 as enquiry has at present gone, have always been present. Hitherto 

 the rods have been considered as if they were ahvays simple and un- 

 branched bodies ; but in many of the Geometrcs, as in Minoa euphorhiata 

 (see figure, post.), where the 8th is a typical abdominal segment, and 

 consists of two separate and nearly equal plates, each outer rod 

 divides into two, one branch going to the corner of the dorsal plate, 

 the other to the corner of the ventral one. 



No part of the investigation has been more puzzling than the 

 muscular system connected with the rods. Actual dissection is next 

 to impossible in the Micros, and in them one has to depend upon more 

 or less accidental preparations ; but in the larger insects it is not so 

 very diiScult to trace out the whole course of the muscles. They 

 appear to be special developments of the ordinary recti muscles of 

 the abdomen, and consist of two opposing sets, which may be called 

 respectively protrudors and retractors, according as they push the 

 parts out or draw them in again. Kemembering that the main purpose 

 of the rods is to secure great length of fibre for the muscles, it is not 

 surprising that it is into the points only of the rods, where they are 

 all crowded together, that the muscles are inserted, and that the shanks, 

 however long they may be, remain free. 



Let us now examine a large insect like TriphcBna pro7iuha, taking 

 the retractors first. 



At the back part of each dorsal plate of the abdomen are a pair 

 of transverse black lines, one on each side the middle line, marking 

 the attachment of flat bands of muscle, the so-called recti, which 

 retract and telescope the segments into each other. On the 7th seg- 

 ment, however, in place of these flat bands, there come off four 

 roundish bundles of muscular fibres, two on each side, which pass 

 downwards, and which are inserted into the points of the four rods, 

 one bundle into each rod Their action is to withdraw both ovipositor 

 and sheath within the body, and from the circumstance of having a 



