CONCLUSIONS 481 



because of their adaptations to their environment that these 

 animals come to resemble one another, and these adaptations 

 have of course no value in determining affinities or descent. 



Another phenomenon may now be considered, which is 

 in some ways intermediate between homology and convergence. 

 It is often the case that in two groups of related animals which 

 have recently diverged the same evolutionary changes take 

 place. This may be called parallelism, and it is illustrated 

 in certain groups of Ungulates such as the Titanotheres and 

 the rhinoceroses. In several distinct stocks of Titanotheres 

 peculiar bony knobs appear on the skull. These structures 

 were not visibly present in the common ancestor of the 

 forms which have evolved them ; the structures cannot there- 

 fore strictly be called homologous, yet they are so similar 

 that it is impossible to avoid the impression that they 

 have some common cause. The independent development of 

 such similar structures in related groups of animals is often 

 ascribed to a so-called process of " Orthogenesis,' ' or variation 

 along " straight " and constant lines. The working of this 

 process in two or more related groups is supposed to result 

 in parallel evolution. 



Now it is worthy of note that when tracing lines of descent 

 through fossil forms, it is rarely possible to identify one form 

 as the direct ancestor of another. Instead, it is more usual 

 to find that one fossil form is related to the ancestor of another, 

 because it possesses characters which that ancestor must have 

 possessed, while at the same time showing other characters 

 which proclaim that it had diverged from that ancestor. The 

 characters of the ancestor in question are, of course, to a certain 

 extent deducible from those of the form descended from it. 



The incompleteness of knowledge of the fossil record 

 makes it difficult to find " fathers," but it supplies a number of 

 " uncles." The question now is this : why do the " fathers " 

 and " uncles " resemble one another ? Cynognathus itself 

 is not the ancestor of the mammals, for in several respects it is 

 too specialised, but it must have evolved parallel with the 

 ancestor of the mammals or it would not possess so many 

 similarities. In the same way it can be shown that the later 



2 1 



