92 EVOLUTION OF LIVING ORGANISMS. 



Petaurus, have become remarkably alike. Very striking 

 also may be the analogous resemblance between two 

 organs of different origin, but fulfilling the same definite 

 function, as the eyes of Polychsete worms, of Molluscs, 

 and of Vertebrates. Mimicry and protective resemblance 

 also are special examples of convergence. Yet however 

 close the resemblance may be, it is generally essentially 

 superficial, and the separate origin of two structures is 

 usually betrayed not only by fundamental differences, 

 but also by innumerable details. Just as the expert can 

 often detect a forged antiquity by careful inspection, so 

 the comparative anatomist distinguishes analogies from 

 true homologies. It is worth insisting upon this subject, 

 because the French philosopher Bergson has recently, we 

 think, greatly exaggerated the extent and somewhat mis- 

 represented the significance of convergence. 



Coming now to classification, it is the sorting out of 

 organisms into groups according to their natural affini- 

 ties. Individuals are grouped into species, species into 

 genera, these again into families, orders, classes, and 

 phyla, divisions of increasing size and importance. 

 Before the doctrine of evolution was accepted, it was 

 thought that these groups had definite limits based on 

 some separately created and fixed unit. At one time it 

 was supposed that the comparatively large genus was the 

 unit of creation, the species and varieties within it be- 

 ing merely fluctuations caused by external influences. 

 Linnaeus maintained that the smaller group, the species, 

 is the originally created unit, and succeeded in establish- 

 ing his views so firmly that they grew into a dogma from 

 which the systematist even of the present day has not 

 entirely freed himself. But the consistent evolutionist 

 recognises that so-called " species " are merely closely 

 allied individuals descended from a common ancestor, 

 which normally interbreed, and are sufficiently alike to 

 be conveniently called by the same name. All sorts 

 of vain attempts have been made to draw up stricter 



