( fix ) 



eacb is so very small, that iieitlicr the Papuan iinr the I'al-iearctie ooniitries can 

 be considered to be a Region in the distribution of the S/j'/in(jida\ But if we 

 sink them both to the rank of a Subregion of the Oriental Region, witli wliich 

 the Palaearctic conutries have 18 genera and Papuasia 19 in common, the most 

 characteristic feature of the temperate Old World would be swallowed up ; the 

 Oriental Regiim would receive foreign elements, and hence the relation of the 

 faunae of the Old and New Worlds be very ranch obscured. The Palaearctic 

 countries being particularly important as a connecting-link between the Old and 

 New Worlds, having many atHnities with North America, as we shall see further 

 on, we give tliem the rank of a Region, in spite of the close relationship with 

 the Oriental famia. Here again we see that the purely numerical consideration 

 of a fauna does not give its true position in respect to geographical distribution. 

 In order to understand the composition of the fauna of a district, the 

 geographical origin of the various members has to be inquired into. A point 

 essential towanls this object is manifestly the knowledge of the relationship of 

 the various species. And as the grouping together of species into genera is 

 meant to be a grouping of species of the same origin, it is further evident that 

 the correct composition of the genera and their correct position in the classification 

 are necessary premisses for sound zoogeograjihical research. E.xponents of zoo- 

 geography who rely on catalogues — in which the distribution is far more erratic 

 than in Nature — having themselves not enough knowledge of the animals to 

 detect mistakes, cannot go beyond the numerical stage iu the treatment of the 

 distribution, and, if they do, will as often be wrong as right in their conclu- 

 sions, constantly finding in the faunae discrepancies and similarities which are 

 nothing but the result of an insufficient study of the animals iu question. If 

 we took, for instance, the classification of the Sphingidne in Kirby's Catalogue, 

 Butler's Ref/sion, Druce's Ileterocera of Central America (iu Biol. Centr. Anier.), 

 and Hampsou's Motlts of India as the basis for the exposition of the geographical 

 distribution of these insects, we should find a far greater similarity between 

 the Neotropical Region and the tropicus of the Old World than there really 

 exists, numerous genera being said to occur iu both tlie Eastern and Western 

 Hemispheres, while closer inspection proves these genera to be quite " unnatural " 

 (heterogeneous), and the various heterogeneous components to belong to different 

 genera ])eculiar either to the New or to the Old World, Protoparce, " Diludia,''' 

 '■'■ Triptogoii,^^ '■' ^[mh/di/x" Pneudosphina-, Daraps'i, Pcrgesa, Macroglossum, 

 '^ Aellnpus" " (7/oerocampa,"' Theretra, etc., etc., being names employed 

 erroneously for mixtures of Old and New World Spluntjidar. 



Proceeding to examine more closely the composition of the iSi)hingid fauna 

 of the various Regions, we take them iu the order as they come iu the tables of 

 distribution on pp. cv. and cvii. The DTeotropical Region is inhal)itedby 33 genera, 

 containing 'ZZ~ species. Two of the genera (Celerio and 1 terse) and one species 

 (Celerio lineata.) are cosmopolitan. The remaining 37 genera may be classed 

 in two groups : 22 which are confined to the Neotropical Region, and 14 which 

 occur also outside it. We will examine the latter first. They are : Protoparce, 



