( cviii ) 



The New 'Wdilil has 54 genera which do not ocenr ontside it ; 100 are 

 peculiar to the Eastern Hemisphere ; while 7 are common to both. The pro- 

 jiortion between genera and species is, therefore, nearly the same in either 

 hemisjihere. Taking info account the origin of the genera as shown in the 

 following i)ages, and leaving aside the truly cosmopolitan Celerio, there are 53 

 genera of American extraction, against 113 of Old World origin. 



Some significant discordance will be found in the numerical distribution of the 

 genera and species when comparing the two tables (pp. cv. and cvii.). The Western 

 Hemisphere has one and a half times as many species of Achfrontiinae as the 

 Old World, but they are more uniform, belonging only to 10 genera; whereas 

 the species of tlie Eastern Hemisphere are grouped in 27 genera, representing 

 many more ste|)s in the development of the Acherontiinae than do the Trans- 

 atlantic forms. On the other liand, the 26 Old AVorld species of Sesiinae belong 

 only to 3 genera, while America possesses 22. In Choevocaminnae. America, with 

 its 3 genera, of whicli 1 is cosmopolitan {Celerio), stands far behind the Old 

 World, which has 12, though the proportion of the species is 53 : 85. 



The diflfereuce in the proportion of the species and genera is instructive in 

 another respect. Looking at the table of genera, it will be manifest tiiat the 

 student of geographical distribution who bases his conclusions on the Sesiimie 

 alone would have only one centre of development, which has given off a few 

 branches to other countries ; tliere would be no cjuestion of "Regions." If he 

 took the Choerocampiiuie as the basis of his zoogeograi)hical division of the 

 earth, Aethiojiia and Indo- Australia might become Regions and the rest of the 

 globe a])pendages of them. If he based his exposition on the Pkilampelinae, 

 Africa, ludo-Australia, and North America would rank as Regions, and South 

 America, Europe, and temperate Asia appear to be /oogeographical appendages of 

 them. While in the case of tlie Arhcrontiiimt'. \heve. would be four Regions. It 

 is evident from this that the division of the globe into /.oogeographical Regions 

 is different according to which group of animals is taken as the sole basis of 

 the division, or, in other words, that the division which is correct in one group 

 (if animals does not necessarily apply to every other group. That the Neotropical, 

 Aethioiiian, and Oriental Regions are nevertheless natural zoogeographical 

 districts in all larger groups of land and iieshwater animals is not to be 

 wondered at, since these Regions comjjrise each the subtropical and tropical 

 parts of a continent and its satellites of islands. America north of Me.xico 

 and Europe and temperate Asia are by no means so well characterised as, and 

 co-ordinate to, the other three Regions. 



In the above tables of distribution we have included the Papuan Spliiiigidae 

 in the Oriental fauna. Australia, New Guinea, the islands farther east, and 

 those westward to the Moluccas have 20 genera of Hawk Moths, of which 6 

 liave not been found elsewhere. The Palaearctic Region has 7 peculiar genera 

 out of 3U. These figures alone would show, therefore, that the Papuan countries 

 were as much entitled to the rank of a Region as the northern temperate Old 

 World zone ; or, on the other hand, since the proportion of genera peculiar to 



