1891.] 179 



the plates, and owe their existence to the necessity of providing that 

 the muscles which are attached to them, and on which the movements 

 of the egg-laying apparatus depend, should have great length of fibre. 

 For the muscles do not act on the principle of a lever, as in the limbs, 

 but pull upon the part directly, producing simple protrusion and re- 

 traction, so that the amount of motion that is possible can be neither 

 more nor less than the amount of shortening the muscles undergo on 

 contraction, in the same way as an object is pulled in, measure for 

 measure, by the shortening of the cord to which it is fastened. Con- 

 sequently, in those cases where the parts have to be protruded freely 

 the rods are proportionately long, while they are comparatively short 

 under the opposite condition of things, and may even be absent 

 altogether when the amount of protrusion is very trifling, or other 

 means are taken for securing length of muscular fibre. Complete 

 absence of the rods is, however, very exceptional. I have never met 

 with it in the case of the inner pair which work the ovipositor, and 

 for various reasons I much doubt whether it ever occurs with them ; 

 but here and there in one or two of the great groups the outer pair 

 are suppressed. In some of the Diurni, as for example Hipparchia 

 Janira and Aiy^/nnis Ewphrosyne, they appear to be absent. The 

 Hipparchia has the 8th segment converted into a ring, whose widest 

 parts are at the sides, where the rods, if present, would have been 

 attached. In the Argynnis the segment consists of a separate dorsal 

 and ventral plate, the latter being narrow in the middle and wide at 

 the sides, which reach back some way. Thus, by the same means, 

 viz., lateral dilatation of the plate or ring, sufficient length for the 

 muscles is obtained in both insects without the intervention of the 

 rods. Thanaos Tages, the only other butterfly I have examined, does, 

 however, possess the rods ; but they are very rudimentary, and are 

 merely short crooked processes, attached one at each corner of the 

 ventral plate. Again, in the genus Cramhus, so far as I have looked 

 into it, the outer rods are also absent. Here the rather narrow sheath 

 is closely connected with the gigantic apparatus concerned in impreg- 

 nation, from which in all pi'obability the muscles get their attachment. 

 Perhaps I ought to say something about this apparatus. In Lepido- 

 ptera it is quite independent of the ovipositor, opening below it and 

 further back ; it is a most complex and remarkable structure, though 

 it is only necessary here to speak of its external portion. This in its 

 simplest form is a small, x-aised, chitinous opening, communicating 

 with the abdomen through the intersegmental membrane uniting seg- 

 ments 7 and S. Frequently it becomes a short canal by being prolonged 



