( Ixxxvi ) 



Hcsides tlic Vil.uiLial vfiitral area there is tlic eiglith tergite wliicli forms 

 jiart of the ciiiuilatoiy apparatus of the female. This tergite, covered by the 

 seveutli, is never spinose ; it varies in size anil shape (Fl. XV. f. 9. 13 ; 

 IM. XVI. f. (i. 9— 11. lO: IM. XVII. f. .■)— s), and is in many cases different in 

 tiie sjiecies of tlie same genus, occasionally even in geographical forms of the 

 same species (PI. XVIIl. f. .5. 8). The most remarkable eighth tergites are 

 those rci)resented by PI. XVII. f. 10 and PI. XVIII. f. 7. In the former there 

 exists ajiparcmtly a gland (o), from which a groove leads anad to a deep cavity. 



Tlio geographical variability of tiie sexual armature is very instructive for 

 tlie beginner in the study of comparative morjihology. The similarity between 

 the structures of the various subspecies is evident in sjnte of the differences ; 

 the modifications are easy to perceive, and there is no difficulty in recognising 

 the homology. The meaning of evolution and of relationship is very clear wiien 

 a series of subspecies is compared with one another and with the nearest allied 

 species. 



No less instructive is the asymmetrical development of the armature found 

 in a uuuiber of Sphingidae. From one jioint of view the asymmetry is even 

 more interesting than the geographical variability. For it is in tlie case of 

 geographical modifications not rarely impossible to say with certainty which of 

 tiie modifications is the oldest, while there can be no doubt about the asymme- 

 trical armature being a development from a symmetrical one. Therefore it is 

 easy to make out by comparison of all the allied species which was the ancestral 

 form of the armature. We liave here not only the line of development, but also 

 the direction followed (see p. l.xxiv.). Apart from tlie penis, the armature of 

 which is nearly always asymmetrical, we meet with asymmetry in the claspers 

 and harpe and in the tenth segment of the male. There is outside tlie Sesiinae 

 only one case of obvious asymmetry in the liarpe, which occnrs in a species of 

 Poli/pti/chus, a genus oi Ambidirinap. In Pohfptyckus dentatus (PI. XXXIV. f. (j) 

 the harpe has a long ventral process {pp) ; the processes of the right and left 

 harpe are the same in length, and both harpes are quite separate from one another. 

 In a clo.-;ely allied species, Poli/pti/chus trilineatus, a remarkable modification 

 is observed. The ventral processes (pr) of the right and left harpe are fused 

 proximally in the Philippine form of trilineatus (PI. XXXIV. f. 7) ; the process 

 of the right harpe (left in figure) is small and slender as compared with the 

 process of the left harpe. In the North Indian form, P. trilineatus itndatns 

 (PI. XXV. f. '1, pr), the two processes are fused together to a fishtail-sliaped 

 structure, ami iu the Ceylonese subspecies, P. trilineatus luteatus (PI. XXXIV. 

 f. lU), the process {pv) appears as a single asymmetrical one, which does not 

 show a trace of its being the product of the fusion of two processes, except at 

 the base, where tlie two halves e.xteud upward on each side, surrounding the 

 penis-sheath. Tiiis combination of geographical variation and asymmetrical 

 development is very interesting. It will be seen, moreover, from the figures 

 quoted, that the upper, shorter, process {pm) of Poli/pti/cltns dentatus is of different 

 length right and left in P. trilineatus (pdl and /jdr). 



