xlv 



Bolivia to the Isthmus of Darien. The species could not volun- 

 tarily pass over, nor by gradual migration along the coast could 

 they well double the end of the chain near the mouths of the Atrato 

 and Magdalena, and so pass to the eastern side, for the Sierra 

 Nevada bars the way. 



Insects, I believe, would offer better data in discussing this 

 question of barriers than almost any other group of land animals, or 

 than plants ; they are more limited in range than the species of 

 birds, afford a much larger body of facts than reptiles, mammals 

 and shells, and are not so much subject to accidental means of 

 transportation as plants. But although manj^ Entomological col- 

 lectors have visited Guayaquil and the Cordillera, we have no 

 published lists and no authentic information about locahties. Mr. 

 Buckley's journey offers us, then, the chance of obtaining the details 

 so much required, as he collected assiduously all the way up from 

 the level of the sea to the edge of the snow, and the same conversely 

 on the opposite side, writing the locality on the envelope of every 

 specimen. 



I am inclined to think that the efficacy of physical or geographical 

 barriers in limiting the distribution of animals and plants has been 

 much over-estimated, and that this circumstance has vitiated much 

 of the reasoning that has been employed in discussing various 

 difficult problems in Natural History. By physical barriers, of 

 course, are meant barriers of the inorganic world, such as a con- 

 tinuous mountain-range with regard to species of the plains, and 

 conversely, a continuous plain with regard to species of the moun- 

 tains (e. g. Parnassius, Erebia, Oreina, Nebria, &c.). The sea is 

 thus a barrier to land- species, a water-shed to fresh-water species, 

 a continuous tract of forest to species of the savannah or steppe, 

 and so on. Barriers of the organic world, which of course are 

 "physical" also, are quite a different set of agencies. They are the 

 hindrances offered to the dissemination of a species by other species 

 already in full possession of the domain, and well-adjusted to its 

 conditions by constitution and habits. To this may be added the 

 limitatio}is to distribution observable without anj^ physical obstacle 

 being perceptible. There are certain classes of facts which seem to 

 me to indicate that these less obvious kinds of barrier are far more 

 effective than those more imposing ones of mountain, desert, sea, 

 and so forth. 



