138 C. B. Davenport. 



trivial or individual variations has had something to do w^ith the 

 origin of species. 



For what are species ? They are assemblages which differ in one 

 or more of their characters to so pronounced a degree as to meet the 

 more or less hazy ideals of the systematist. Now it is quite cer- 

 tain that such differences may arise in several ways. I propose 

 to present certain evidence indicating that the differences between 

 certain recognized species are of the order of accumulated individ- 

 ual variations and not of the order of mutations. Such evidence 

 will then go to prove that evolution may, in some cases, take place 

 without mutation. 



Evidence as to the method of origin of species should be looked 

 for in a group where closely related species are found to-day. 

 Any interspace of distribution or time should be searched to see 

 whether there are any graduated, transitional forms, or whether, 

 on the contrary, a sudden change of character occurs. There are, 

 in general, two kinds of series available for examination: one is 

 the geographical, the other the palaeontological series. At the 

 present day two "species" are often found occupying two more or 

 less distant regions of the continent. If we examine the inter- 

 vening territory shall we find all gradations between the species or 

 shall we find a sudden transition from one to the other ^ We 

 should expect the sudden transition only if mutation is the sole 

 method of origin of species; on the other hand, if fine intergrada- 

 tions appear these would speak for evolution by trivial variation. 

 So, too, if a series of fossils connecting one species A with a second 

 B are examined a sudden transition or a gradual one will be found 

 as mutation has or has not acted in the case in question. 



The facts of geographical variation are well illustrated in the 

 broad territory of North America. Our eastern song sparrow, 

 to cite a single example, was formerly thought to be replaced by a 

 distinct species on the Pacific Coast in the vicinity of San Fran- 

 cisco, by a small light-colored desert species in Arizona and by a 

 large dark species in Alaska. Later collections from inter- 

 mediate localities revealed intermediate forms so that the different 

 geographical forms are now regarded as varieties of one wide- 

 spread species showing fine gradations in coloration and structure 

 from place to place. In speaking of the results of the study of 

 geographical variation of land tirds in America, Newton ('93-'96, 

 p. 343) says: "The great fact was established that, given a species 



