Barbets, and their Distribution. 245 



belong to different families. For example, humming-birds 

 (Trochilidce) are very numerous all over the Neotropical 

 Eegion — under which name are comprised Central and Southern 

 America — but are quite unknown in Asia and Africa. In the 

 same category we may place the toucans (BJiamphastidce) , Jaca- 

 niars (Gcdbulidce) , motmots (Momotidce), puff-birds (Biicconidce), 

 curassows (Gracidce), tinamous (Tinamidce) , and numerous other 

 families which are more or less generally distributed throughout 

 Central and Southern America, but are not found in the tropics 

 of either Africa or Asia. In the same way, the bee-eaters 

 (MerOjpidce) , hornbills (Bucerotidce), rollers (Goraciidce), sugar- 

 birds (Nectariniidce), honey-guides (Indicator idee) , orioles (Orio- 

 lidce), and other families, are characteristic groups of the orni- 

 thology of the Old-World tropics, but are not known to 

 exist in the New World. The cases where members of the 

 same natural family of birds are met with in the tropics of 

 both hemispheres are mostly those in which the families are of 

 extended geographical distribution. For example, the falcons 

 (Falconidce) , owls (Strigidce), swifts (Gypselidai), swallows (Hi- 

 rundinidce) , and thrushes (Turdidce), are widely diffused families, 

 which are represented in every part of the earth's surface, and 

 are abundant in the tropics of both the Old and the New 

 World. In the same category may be placed most of the 

 families of waders and water-birds, which are usually of very 

 wide range — in some few cases even the same species occurring 

 in every part of the world. But it is quite an exceptional case 

 for a family of birds confined to the tropics to be found in both 

 hemispheres, and besides the Barbets, the only well-defined 

 families presenting the same phenomenon of distribution are 

 the parrots (Fsittacidce) and the trogons (Trogonidce) . Mem- 

 bers of both of these groups, as in the case of the Barbefcs, are 

 found in the tropics of Africa and Asia as well as in those of 

 America. 



Now, assuming the derivative origin of species — that is, 

 the descent of the various members of natural groups (whether, 

 we call them genera or families) from a common ancestor at a 

 more or less distant epoch, let us see what deductions may be 

 derived from these facts of geographical distribution. 



In the first place, looking to the general diversity of the 

 avi-faunas of the tropics of the two hemispheres, it is evident 

 that the lands which constitute their seats must have remained, 

 as they now are, for a long period of ages, separated by an 

 extensive barrier of ocean. Upon no other hypothesis than 

 this can the number of extensive groups peculiar to each of 

 them, and the general dissimilarity of the two avi-faunas, be 

 explained. 



But on the other hand there must have been a time when 



