( cxxxiv ) 



cosmoiiolitau ones (/frn<e and Celrrio). The affinity which there exists between 

 the Neotro])ic:al and Aethiopian Kegions is shared by India. 



Since low types of all tlie subfamilies of >Sphiitgidae, with the exception of 

 the Sesiinac, occur in the tropics of the Western and Eastern Hemisi)heres, there 

 is indeed a close affinity in this respect between the tropics of the Old and 

 New Worlds. But the affinity constituted by the presence of allied low types has 

 quite a difl'erent bearing on (|uestions of ojeographical distribution from that affinity 

 which is established by the occurrence of closely related specialised genera and 

 species. Let us take, for instance, the Acheroiitiinaf and Amhidicinae. The 

 lowest African, Indian, and American geneia Xanthopan, Meganoton, Cocijthi», 

 and Protoparcr are as nearly related to each other as the younger derivatives 

 llyloicua, Ceratomia, Thamnoecha, etc. ; and the generalised Ambulicine genera 

 Frotumbiih/x, AmjAypteras, Compsogene, and Oxi/ambuh/x are phylogenetically as 

 closely connected as are the specialised genera Sphinx, Amorpha, Pachi/sphinx, 

 Phi/Uosphingia, Cressonia, etc. The migration northwards and westwards into 

 the Nearctic Region of the derivations from the ancestral Oriental A?nbtdicinae, 

 represented by Amplypterus in the present epoch, and the extension of the 

 derivations from the lowest Neotropical Acheroitiinae, Protoparce, or an extinct 

 genus similar to it, north and eastward into the Palaearctic Region as far south 

 as the Himalayas, and, further, the modification of either branch into a number 

 of genera, have required a long time. Since not one of the younger derivations 

 from the Oriental Ambulicinae and African Acherontiinae occurs in South 

 America, and none of the younger developments of the Neotropical branches of 

 these subfamilies are found in Africa or India, thougli the younger developments 

 have become distributed northward as widely as stated above, it is obvious that 

 there was no direct connection between the Neotropical Region and Africa and 

 India during that period in which the evolution and distribution of the younger 

 members of the two subfamilies took place. That is to say, the time which 

 has elajised, since the African Xanthopaii and the Neotrojiical ('oa/tius, or the 

 Indian Compsogene and the Neotropical Ampli/pterus, closely allied though they 

 are, became geographically separated, is more remote than that which elapsed 

 since the Nearctic and Palaearctic Regions exchanged a number of specialised 

 genera. The distribution of the other subfamilies confirms this conclusion. The 

 generalised Sesiinae occur only in the Neotropical Region, not one species being 

 peculiar to any other region, though one single specialised— the most highly 

 specialised — branch has crossed to the Palaearctic Region, and established itself 

 in two genera in the troj)ics of the Old World {Sataspes, L'ejthonodes). The 

 oldest genus of J'hilampelinac is again Neotropical {i'holns). It is sharply 

 separated from the rest of the subfamily, and the only one occurring in the 

 Neotropical Region, besides a derivative fmiu I'l/olas found on the Sandwich 

 Islands {'rinostoma). All the other Pliilampclhidv are of Mid World extraction, 

 and have parti}' migrated from the Eastern Ilemispheie into the Nearctic 

 Region. In both the Seaiinac. and I'liildtupdiwie the gi;ogra])hical barrier 

 between the Eastern and AVestern tropics must have been etfective at a very 



