( xoi ) 



development load iiKlciieudfiitly U< an identical specialisation in the two divergent 

 branches. The sjiination of the abdomen furnishes a noteworthy illustration. 

 In the higher .Scgiii/nr tlic spines become very stronjily chitinised on the tergites 

 and sternites, and those of the i)roximal vow assume a short and broad form in 

 Si\sia, Cephoitfldi's, and Ilaemorrliagia ; the same form of sj)ination is found again 

 in the highest genera of the tribe Xeplidirne of the subfamily rhilamju'linae— 

 namely, in ^facl■oqlosxum, Kliopalopsyche, and Lctcostrophus — and only here. In 

 the Pliilampelinae. there appear here and there characters which are met with 

 again among the Sesiinae — for instance, the very strongly clubbed antennae, the 

 annulate mid- and hindcoxal merum, the close connection between abdomen and 

 thorax, the fan-tail, etc. : while, on the other hand, Philnmpdtnae. have also 

 arrived at similar specialisations as Clioerocampinae — for instance, in the shape 

 of the chrysalis, the sexnal armature, the end-segment of the antennae, and the 

 ocellated spots of the caterpillar. The American Choerooampine genus Phanoxi/la 

 is a derivation from Xi/Ioplianes, from which it has become different in acquiring 

 an ajncal tuft to inner surface of the second palpal segment, a tuft which is 

 present in many Old World Clioerocumjiinae. The peculiar cavity at the end of 

 the first segment of the palpus, on the outer side, occurring in all species of the 

 (.'hoerocampina genus Tlirrdra and the derivations from this genus, is acquired 

 also by a few species of the allied genus Jlijijintioi/, and it is highly surprising 

 that the same peculiar sjiecialisation, which does not seem to occur outside the 

 Sphingidae, is as cons])icuonsly developed in one solitary species of Acherontiinae, 

 in Megacorma oblit/iia, the relationship of which with the Clioerocampinae is 

 very distant. This similarity reminds one of the projecting tongue-case of the 

 pupa of the (^'hoerocampiue- genus Uliyncliolaba and many Acherontiinae, a 

 Sphingid character confined to that one genus and the one subfamily. The 

 pattern of the Hawk Moths offers also obvious illustrations of the kind of 

 resemblance under discussion. The most striking is the similarity in the 

 pattern of the abdomen of the lower Acherontiinae and the higher Philainpelinae. 

 Here we find those conspicuous yellow side-patches bordered by black which are 

 restricted to these insects. The white abdominal belt of the American genus 

 Sesia and the African Philampeline genus Lciwoxtrophiis, which belt Entomo- 

 logists have erroneously considered to indicate a close relationship between the 

 insects, may also be mentioned, as well as the black-bordered yellow hindwing 

 occurring in many Uphinyiilae and other Moths, and the white fringe of the 

 first abdominal tergite of the Oriental Tieilephila hi/pothous and the American 

 Aleiiron iphis and allies. The antemedian and discal blackish brown Viands on 

 the forewing acquired here and there in Sphingidae are the resnlt of the inter- 

 spaces being filled in with blackish brown scaling. These examples of similarity 

 are a kind of belated expression of relationship. 



It will be observed that in a few of the instances mentioned the agreement 

 in some character affects the aspect of the insects which have the character in 

 common. There is a complete gradation from similarities in some small detail 

 to similarities of the ensemble. 



