102 P^om Magazines, S-c; [,sf o"ct. 



" It is sufficiently obvious," the author states, " that the avi- 

 fauna of a given country is distributed neither accidentally nor 

 uniformly ; its dispersal is, on the contrary, both definite and 

 uneven. Despite their exceptional mobility, birds are essentially 

 local ; the ' universally distributed ' species is at most universally 

 distributed only within the limits of a uniform environment. In 

 one sense, and from one aspect, the environment may be said to 

 determine the avifauna. The locahzation of birds is directly 

 related to the locahzation of environmental types. Since this is 

 so, an accordant classification of the phenomena of distribution 

 seems to be necessary to a full illustration of the whole subject. 

 But the term environment stands for a complex of varied and 

 intricate factors, with differentiated reference to the fauna, and 

 the preliminary difficulties arise in the search for the dominant 

 or immediate agents in dispersal. The local existence of a bird 

 as a breeding species is dependent upon the local conjunction of 

 certain environmental units ; and distribution is uniform in pro- 

 portion as these essentials are uniformly and evenly dispersed. 

 A shortage in one direction is not balanced by superfluity in 

 another. The specific range is necessarily contained within the 

 region of efficient food supply ; but it is of equal necessity con- 

 tained within that of suitable climatic conditions, and within 

 that of nesting-ground. A sufficiency of food and the necessary 

 climatic conditions do not permit existence if nesting-ground is 

 absent ; an abundance of food and nesting-ground is useless if 

 the meteorological conditions are inimicable. Where the 

 requisite conditions are unequally distributed in space, or 

 relatively unequal in supply, there the dispersal of a species will 

 be correspondingly uneven, limited by the most local, or the least 

 adequately supplied, of the essentials. In the case of many birds, 

 an increased supply of nest-sites is alone sufficient to allow of an 

 increase in population, which implies that the possible maximum 

 is normally not attained through a relative shortage of nest-sites. 

 In such instances, distributional range is co-extensive less with 

 food-supply than with nest-sites. The latter is, therefore, so 

 far the limiting factor in distribution. If a common limiting 

 factor exists, and can be recognized, then on it a classification of 

 the facts of distribution might be conveniently based. But this 

 can hardly be looked for ; a varjdng supply of varying needs must 

 be expressed by a limiting agent differentiated according to species. 

 Nevertheless, it appears probable that this differentiation is 

 confined within relatively narrow limits. The governing or 

 limiting factors in bird distribution, apparent in part through 

 direct observation and comparison, are perhaps thrown into 

 clearest relief in the phenomena of range-extension — which process 

 may be graded separately in degree as it represents predominantly 

 either (i) the colonization of new areas ecologically homogeneous 

 with those of origin, or (2) extension of range involving adjust- 

 ment to new environmental conditions. Theoretically, extension 

 may be due to such features as numerical increase, or to environ- 



