OTHER THEORIES OF SPECIES-FORMING. 257 



between Melospisa cinerea montana and Melospisa cinerea rufina, 

 between which there is continuous distribution and free interoscula- 

 tion. But we cannot expect any two species of birds or other 

 animals to present the same degrees of differentiation in the same 

 length of time or under the same conditions, much less under 

 different conditions. For in no two animals is the physical organ- 

 isation, in all respects, exactly the same. 



"In a given aggregation of individuals constituting a new colony, 

 a certain amount of time is necessary for the set of environmental 

 factors to become operative in bringing about new inheritable 

 characters to a degree perceptible to us. Then the inherited effects 

 of invasion and cross-breeding from season to season from the 

 adjacent parent centre of differentiation will be evidenced less and 

 less, as time elapses, as the distance from this centre increases. 

 The offspring of successively further removed unions will, of course, 

 inherit to a less and less degree the distinctive characters of the 

 ancestral stock on one side and more and more of the incipient 

 ones on the other. 



"If, now, the distance is great enough to permit of the time 

 required for adaptive manifestations to become innate, then we 

 would find new characters making their appearance distally nearest 

 the new centre of differentiation. If the distance were too short 

 we would not find new characters showing themselves because 

 they would be constantly crowded down by the influx of the old. 

 The time factor may, therefore, be reduced by the intervention of 

 an impassable barrier. As an instance, we find three (and there 

 are probably two other) insular forms of the song sparrow within 

 a limited distance among the Santa Barbara Islands, while through 

 the same distance on the adjacent mainland there is but one. Or, 

 in the case of continuous distribution, the time element may be 

 comparatively lessened by the great distance between the range 

 limits, and it may be still further decreased as these limits lie in 

 faunal areas of more emphatically different nature. The horned 

 larks, as well as song sparrows, furnish us several good examples 

 of the latter two rules. 



"It is isolation, either by barriers or by sufficient distance to 

 more than counterbalance inheritance from the opposite type, that 

 seems to me to be the absolutely essential condition for the differ- 

 entiation of two species, at least in birds. 



"A strong argument in support of this conviction is that we 

 never find two 'sub-species' breeding in the same faunal area, and 

 no two closely similar species, except as can be plainly accounted 

 for by the invasion of one of them from a separate centre of 

 differentiation in an adjacent faunal area. An appropriate instance 



