Tabular Review. 57 



strong but not exact, and is satisfied with a general imitation, not in- 

 sisting on all minute details. Or, to express it more scientifically, the 

 law of inheritance tends to, and under like conditions of environment 

 always will, produce exact copies, but because the like conditions of en- 

 vironment never do, and, in the nature of things, never can exist, there- 

 fore no generation is, or ever can, be an exact copy of its predecessors. 

 Natural selection intervening, designates whether any modification thus 

 made shall be perpetuated and intensified by further accretions in the 

 same direction, or be suppressed. The table to be complete should 

 contain a million terms instead of twenty-two. It is impracticable, of 

 course, to construct such a table, and probably always will be. The 

 stages represented in the table are to be regarded as the centers of wide 

 districts, the borders of which are not sharp boundary lines, but which 

 fade from one to another by so gradual a process that no boundary can 

 be fixed. If we had all the inhabitants of a village collected in a body, 

 and should then try to select out, by means of their family resemblances, 

 those related to each other, we should undoubtedly get things mixed. 

 If the population were of different races, we could easily enough pick 

 out the full-blooded Africans from the full-blooded whites. The table 

 is constructed something on this plan. The great and obvious succes- 

 sive terms are here; but the filling, which makes one life to extend gen- 

 eration after generation, individual by individual, from the monera and 

 away beyond for he is not the beginning of life down to the brainy 

 Caucassian, is all left out. The column on the right shows the types 

 now living that most resemble the ancestral stages in the opposite col- 

 umn. It does not follow from this that the tribes on the right have 

 stood still ever since our ancestral line was equivalent to what theirs is 

 now. But supposing them to be subject to modification as we are, it 

 follows that their ancestors were also ours, say ten million years ago, 

 and a far lower type than they are now, but that our branch of the 

 family reached the development in say five million years, the equivalent 

 of which the other branch has taken the whole ten million for. This 

 very obvious reflection amply accounts for the fact that some stages, 

 which our line must have passed through, have no exact counterpart 

 now living. Thus the fifteenth stage was the first race that developed 

 the amnion sac. And we feel morally certain that from such a race all 

 the reptiles, birds and mammals have descended, diverging from each 

 other through lapse of time and changes of environment. But none of 

 the descendants now living are exactly like that common ancestor. It 

 is no wonder. Philologists assure us that the long list of languages named 

 the Indo-European, and embracing as general divisions, the Celtic, 

 Greek, Latin, Sanscrit, Persian, German, Sclavonic and Scandinavian, 

 and whose sub-divisions run into hundreds of dialects, and are spoken 



