( civ ) 



while the true cause of tbe absence or scariity of those insects in the Malay 

 Archipelago is jirobably of (luite a different kind, as we shall see later on. 

 The North American Spliccoiliiw ubijotii is confined to the Atlantic side of the 

 Continent : the only other species of the genns (the only near ally, in fact, of 

 abbotti) is restricted to the Pacific side of Asia (Amurlaud, China). The Indian 

 geaera Apocah/iisis and l'si'U(loihilbiir:t stand quite isolated in the Old World, 

 their nearest (and very close ally) being the Neotrojncal I'^nrifc^lnttis. The 

 South African genus lllioiliifra comes nearest to the Falaearctic Pfn/r.s^. If we 

 compare such cases with the distribution of Cderio, of which one species is 

 cosmopolitan, one Holarctic, several Palacarctic, and one each Malagassic, Argen- 

 tinian, ('hilian, and Cuban, while two are confined to the Sandwich Islands, it 

 becomes obvious that the interrupted range of genera or allied genera is the 

 result of the extinction of the iusects (or allied species) in the now non-inhabited 

 intermediate districts rather than the result of great geographical changes. We 

 have to do here with the remnants of once more uniformly distributed groups, 

 remnants existing where by chance the conditions were favourable for their 

 preservation. However, it is not our intention to speculate on the apparently 

 abnormal distribution of single species and genera, fascinating as it is, but to 

 lay before the reader a short summary of the actual stale of the distribution 

 of the tijiliinyiilue, from which he will better understand the composition of 

 the Sphingid fauna ot tlie various areas than from a speculative elucidation of 

 single cases. Accepting the Wallacean division of the globe into five Regions, 

 we have a distribution of the 770 species of Hawk Moths, as illustrated in the 

 table on the opposite page. 



There are only 2 species common to the Western and Eastern Hemispheres, 

 each represented in America by 1, and in the Old World one by 2 and the 

 other by I subspecies. Of the remaining 768 species, as yet known, 293 are 

 peculiar to the New World, includinK the Sandwich Islands (3 species), and 475 

 to the Eastern Hemisphere. Since the Papuan Subregion and Africa harbour 

 without doubt a great many undiscovered species, proportionally more than South 

 and Central America, we can estimate the proportion of the sj)ecies inhabiting the 

 New and Old Worlds as being 1 : 2, or presumably 32U-odd to 650-odd. The 

 snj)eriority of the East over the West of the globe in the number of species of 

 Hawk Moths is quite intelligible, since tropical America is far smaller than 

 the tropics of the Eastern Hemisphere, does not offer so much variety in the 

 conditions of life, and is not so broken pliysiographically. A heteromorphic 

 country has a heteromorphic fauna. 



The paucity of the SpliiiKjidai' in the temperate Regions is very apparent 

 in the table. North America, Eurojie, North and Central Asia, and Falaearctic 

 Jajian, where few new species will be discovered, if any, possess only lOU-odd 

 species peculiar to them, as compared with 620-odd peculiar to the only 

 partially explored trojiical and subtropical countries. New discoveries will alter 

 the j)roportion so much to the disadvantage of the northern temperate Regions 

 that the proportion will ultimately be about 1 : t), according to our calculations. 



