PHENOMENA OF REGRESSION AND CONVERGENCE 227 



we not prevented by the enormous difference in 

 size which separates them. 



These frequent convergences of the dermic 

 plates or of the teeth of Vertebrates appear to me 

 to be easily enough explained by the fact that 

 Nature cannot indefinitely vary her processes for 

 incrusting the skin of an animal with osseous 

 tissues, or for grouping the points, primitively 

 distinct, which adorn the surface of the crown of 

 a molar. Let these primitive denticles remain 

 conical and apart, and we have the bunodont type. 

 Let them be welded to each other two by 

 two, closing in from front to back, and we are 

 at the lophiodont type ; finally, let these same 

 denticles arrange themselves in V-shaped curves, 

 and we arrive at the semi-crescent or selenodont 

 type which characterizes the molars of the Ru- 

 minants and of several other groups of mammals. 



It is for a like reason i.e. the poverty of natural 

 processes of structure or of ornamentation in very 

 simple organs that the numerous facts of con- 

 vergence observed in invertebrate animals, and 

 especially in the shells of molluscs, are justified. 

 Very curious examples can be instanced : among 

 the Foraminifera, in which the fundamental division 

 into orders is founded on the nature, perforated 

 or non-perforated, of the calcareous test, the 

 Cretacean and Tertiary Alveolines recall in an 

 astonishing manner the Carboniferous Fusulines, 

 either by their exterior form of a lengthened spindle, 

 or by the close coiling of the spiral whorls. Similarly, 

 certain simple Polyps of the Apores group, such as 

 the Turbinolia, would be distinguished with difficulty 



