( xcv ) 



are not necessarily dependent on one another. >Spinosity of tlie tibiae, lor 

 instance, occurs in very mauy insects without a reduction of tlie tiliia liaving 

 taken place, and eye-lashes are not ahvays a sign of a reduced liead. A 

 conspicuous specialisation which is of common occurrence in insects is the pro- 

 longation of the foretibia into a pointed thorn. We find this thorn or claw 

 among all subfamilies of Sjiliiiitjidnc except Choerorampinae. The i)rolonged 

 outer spines of the first i)rotarsal segment (generally three in number) found in 

 many Sjihinyidae and other Heterocera are also a specialisation which does not 

 necessarily indicate relationshi]), and so are the clubbed antennae of SphimfKlaii, 

 Butterflies, Acgeriidae, some Noctuidae and (ieoiwtridae, Agaristidaf, etc. The 

 pectination of the antennae is of special interest. We find such antennae in 

 several orders of insects. The remarkable point is that there are antennae which 

 are similar ia aspect to the pectinated ones, but are not jiectinatrd. They are 

 of two kinds. The one type is that commonly found in the males of Spliingidae 

 where the seriated cilae look in a dorsal view like lateral e.Kpansious of the 

 segments; and the other type, not met with among Hawk Moths, possesses a 

 heavy subdorsal bristle on each side, which Entomologists have often enough 

 confounded with a true process of the segment. Here we have an obvious 

 resemblance not based on homology. Likewise, the pectination of the antennae 

 of Cressonia and Ceridia, or of Siiturniidae and Eiipterotidac, though the 

 antennae are similar to one another in aspect, are not homologous. Tibiae 

 mottled with single dark scales have very often the appearance of being spinose. 

 In the place of the before-mentioned foretibia! claw there is in some other 

 Sphingidae (for instance, some Poli/ti/chus) a heavy spine homologous to a hair, 

 not to the claw. The horn-like projection of the palpus iu the Acherontiine 

 genera Cocgtius (America) and Codonia (Africa) is the third palpal segment, 

 while the equally conspicuous {)rojection found in the Sesiine genera Ale/iron 

 and E/igo (America) is a process of the second segment. The same segment has 

 acquired a j)rojection similar to the latter in the very distantly related African 

 genus Hi/paedalia. Similarities analogous to the above occur also in pattern. 

 Lines, streaks, and spots on the wings, and spots and belts on the abdomen, 

 superficially but strikingly alike in different insects, often prove on closer study 

 not to be homologous, the belts being, for instance, basal to the segments in one 

 Insect and apical in anotlier, and the wing-streaks standing here upon tlio veins 

 and there between them. Such similarities, which mostly do not affect the 

 general habitus of the insect, are nevertheless very instrnctive. 



There is another kind of equally remarkable resemblance in some details 

 met with among Sphingidae which appears at the end of more or less distantly 

 related branches. It is a priori conceivable that a certain cliaracter of structure 

 or jiattern present in closely related forms is i)reserved when these forms 

 develop divergently further and further. Now, just as this certain character may 

 be lost in the one branch, or in some members of it, at one period, and in tiie 

 other branch at the same or another time, so the character may at one time 

 or the other develop progressively in either branch, and this progressive 



