( xcv ) 



are not necessarily deiiendent, on one another. Spinosity of the tibiae, tur 

 instance, occurs in very many insects without a rednction of the tihia Imving 

 taken place, and eye-laslies are not always a sign of a reduced head. A 

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

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

 among all subfamilies of Sphingidae except Choerocanijnnac. The i)rolonged 

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

 many SphiiKjiclae and otlier Heterocera are also a s[)ecialisation which does not 

 necessarily indicate relationship, and to are tlie clubbed antennae of Sphinyiilae, 

 Butterflies, Argeriidae, some Noctuidar and (reonvtridae, Agaristidae, etc. The 

 ])ectination of the antennae is of special interest. AVe find such antennae in 

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

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

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

 where the seriated cilae look in a dorsal view like lateral expansions 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 (.'ressonia and Ceridia, or of Suturniidae and Enpterotidae, 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 foretibial claw there is in some other 

 Sphingidae (for instance, some Polgti/chus) a heavy s])ine homologous to a hair, 

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

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

 while the equally conspicuous projection found in the Sesiine genera Aleuron 

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

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

 genus Ht/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 another, and the wing-streaks standing here upon the veins 

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

 general habitus of the insect, are nevertheless very instructive. 



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 ciiaracter of structure 

 or pattern present in closely related forms is preserved 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 the 

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

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



