352 PRINCIPLES OF ANIMAL HlOLOdY 



from distribution of present-day organisms over the earth, and from 

 observation of the process. The rather complete series of fossil animals 

 leading up to modern horses and elephants, and the series of cephalopods 

 ending with extinction, as described in the preceding chapter, need no 

 comment as indications of evolution. To them may be added an 

 immense amount of less complete data of fossils, all of which point to 

 the same conclusion, namely, that species and larger groups of animals 

 and plants have changed. Geographic distribution, as repeatedly shown 

 in Chap. 21, likewise requires the assumption of evolution 

 to be intelligible. It should not be necessary to comment 

 further upon it here. The observational evidence will 

 be referred to later. 



Evolution a Change of Species. — Though evolution has 

 effected a separation of groups of high rank (orders, classes, 

 phyla) from one another, it has accomplished this result 

 entirely by modification of species. There is no such thing 

 as single wide cleavages that at once produce cA^en families 

 or genera out of single common stocks. The divergence is 

 everywhere a slow accumulation of small differences such 

 as characterize species or varieties. When life originated, , , '"^' ,■. 



_ _ o 7 Adult free-hv- 



assuming that it did so only once, there was at first only ing barnacle of 

 one species of organism. When a change occurred in a part ^^^ ^''•ti'* t^ u 

 of this group, all experience indicates that the difference of its shell re- 

 could have been no greater than that now existing between moved. 

 species — or more probably varieties. When further changes occurred, it 

 is not likely that altogether the same changes took place in both varieties, 

 so that each of them gradually bi-oke up into two unlike sets of varieties 

 or species. The two varieties produced by the first modification may 

 thus have given rise to two species, later to two genera. By fiu'ther 

 change of species, each group of species pursuing a course somewhat 

 different from the other, these two genera may be supposed to have been 

 transformed into families. Still further changes in species within the 

 families shovdd have resulted in the degree of difference now held appi'o- 

 jjriate to orders. By continued change of species, the orders may have 

 diverged from one another enough to be regarded as classes and finally 

 to have attained the rank of phyla. The coiu'se of evolution has been 

 not to create phyla and then to proceed to split tliem up into groups of 

 lower ranks, ending in species and varieties; it has ratlier gone in the 

 opposite direction, beginning with species and by repeated ciianges of 

 species gi-adually converting them into groups of higher rank. The pro- 

 blem of evolution thus becomes that of the origin of species. 



The Nature of Species. To understand evolution it is necessary, 

 therefore, to know how species arc constituted. A species may be 



