( eii ) 



geogriiiihiciil liuriiLTs becouu' ineffective, and the inflnence of isolation (se<;'i-ega- 

 tiou) is annulled. The wandering Hawk Motlis, like Acherontia atropos, IJerse 

 coi/eoln/// and cingidata, Ilijjpotion cclerio and osiris, Deilephila nerii, Theretra 

 nessus, etc., do not vary geograj)liiciilly, or only in a slight degree. However, 

 considering what we have said alioiit the jiower of flight and isolation, it 

 cannot be wondered at that a good many species of swift-flying genera iiave 

 developed into geograjdiical varieties (= subspecies) in spite of their power 

 of flight. Notwithstanding tliis fact, in the Spliiin/iclae the rule holds good 

 tiiat, on the wlioh', the sluggish sjiecies are more often and more distinctlj' 

 varialde geographically than the swift-fliers, and that, further, the areas of 

 the geographical races of slow-flyers are mostly smaller than those of the 

 subspecies of less slnggisji Hawk Moths, as exemplified by AmbuliciiKte and 

 Clioerocavipitiiw. 



There is only one cosmopolitan Sphingid, Celerio lineuta \ it consists of 

 three subspecies, one inhabiting the AVestern Hemisphere and two the Eastern. 

 Herse conwlvuli and Ilippotioii celerio occupy practically the whole Old World, 

 while Herse clmiulata and Protoparce sexto, occur from Canada to Patagonia. 

 One species {Celei-io gallii) is common to the Nearctic and Palaearctic Regions. 

 Ce.phonodes In/las is found, in three subspecies, from Sierra lieone to Mada- 

 gascar and eastwards to Australia, with a wide gap in the Malay Archipelago. 

 DeilepJrila nerii inliabits the wiiole Aethiopian Region, and extends far into 

 Europe and south-eastwards into AVestern India and Ceylon. A number of 

 species occuir all over the Aethoi)ian Region without being split up into 

 subspecies {Ilippotion eso/i : Busiotliia medea ; Coelonia faU-iiwtdtu : etc.); 

 others are of wide distribution in tlie Palaearctic or Oriental Regions (Macro- 

 glossum utelldtdrnm : Hippotion boerliaviae, velox ; Theretra ?iessus ; Cephonodes 

 picus ; etc.); and others again occur all over the Neotropical Region or go 

 even far into the northern Iialf of tlie New World (.\///oj>//nnes tersa, pliito : 

 Psei(dospkiii.r trtrio : ]Crinrri/is (dope ; Kpistor //fc/id/r/.s : etc.). The distribution 

 of such common species is very instructive in one respect. Although their 

 range is wide, it is nevertheless restricted, and we ask ourselves, What 

 antagonistic factors are there preventing these Sjihinfiidne from occupying the 

 whole globe ? 



A country must be accessible to a sjiecies. If tiiere is a practically 

 unsurmountable physiogra{ihicaI barrier, the species have little chance of passing 

 over it. The swift-flying Spkingidue are not able to cross the Atlantic and 

 Pacific Oceans, none of the New and Old World species being identical, not 

 even Celerio lineata and gallii being the same in the Western and Eastern 

 Hemispheres ; and the Indian Ocean is an equally effective barrier. But there 

 is no such barrier to prevent Macroglossmn stellatarum, which ranges from 

 Morocco and Ireland to Japan and North West India, from going farther south 

 and east in India ; no physiographical chasm faces in Spain the Aethiopian 

 ihppotioii osiris, which extends from Madagascar to the Pyrenean Peninsula ; 

 no geographical barrier keeps the African Deilephila nerii back from settling 



