ORIGIN OF SPECIES 95 



which they severally stopped. What explanation could the 

 teacher of fifty years ago give of the gill-slits, or tail, or a 

 hundred other resemblances to lower vertebrates which the 

 human embryo presents in the course of its development ? 

 They are by no means necessary preparations for adult 

 structure. They never can be useful. Not infrequently 

 they are mischievous. Indeed Man's organs reach their 

 permanent form by many a roundabout road. These 

 digressions are indications of the tenacity of Nature's 

 memory. She can attain her goal only by tracing over 

 again with a jump here and a short cut there it may be, 

 but without letting go of the clue the path which she 

 followed when she first discovered it. 



We are now in a position to understand the influence 

 which Darwin's theorem has had upon the taxonomist. It 

 is no longer enough that he should classify living things 

 according to their natural affinities he must group them 

 according to their proximity to one another on the ancestral 

 tree. His classes are the several stems of this tree, his 

 orders its main branches. Its small branches are genera 

 and its twigs species. Their "natural affinities" do, of 

 course, indicate relationship, but the taxonomist must be- 

 ware of mere resemblances. He can only be sure that he 

 has traced their pedigree when he finds two extant forms 

 uniting losing their differences in a fossil ancestor. The 

 geological record is, however, so imperfect that it is but 

 seldom that certainty can be claimed. 



In any attempt at classifying animals a great and hitherto 

 impassable gap is found between invertebrates and verte- 

 brates. There is, as it were, a wedge-shaped blank in the 

 picture of the ancestral tree. Evidently a vast number of 

 intermediate forms have died out, leaving, according to the 

 common reading of the rock-record, no trace behind. 

 When the highest of the invertebrates of the epoch at which 

 the change occurred began to assume what is now known 

 as the vertebrate type, its transitional form cannot have 



