3o6 TADPOLES OF THE SEA 



that all the existing Ascidians had lost (" suppressed " as 

 the phrase is) the tadpole stage as most of them have, 

 then we should know nothing, or very little, of their affinity 

 to vertebrates. It is, as it were, only by a lucky chance 

 that some Ascidians still begin their life by going through 

 a tadpole stage, and so betray their vertebrate nature. 

 Hence we have to remember that any small and simple 

 sort of animal may possibly be a simplified product of 

 more perfect and active ancestors, although it shows no 

 early stage of growth of a reminiscent character. And 

 we have to seek for evidences of " specialism " in its habits 

 and corresponding structure which would justify the sup- 

 position that it is not really " primitive" in its character, but 

 simplified by excessive development of one set of organs in 

 correspondence with special food supply or means of safety. 

 The notion that the course of evolution or ancestral 

 development, whether of the structure of animals and 

 plants or of the capacities, institutions, and activities of 

 human races, may often have pursued, not a continuous 

 progressive line of improvement, but a relatively back- 

 ward movement of simplification and reduction, is of 

 great philosopic importance. Many savage races are, no 

 doubt, more primitive than those now civilised, and pre- 

 serve some of the characters of primitive man. But some 

 savage races appear to be the descendants of ancestors 

 who had attained to a relatively high degree of civilisation, 

 which the existing stock has lost, and are examples of 

 the same process of simplification (often called " degenera- 

 tion "), as that which we see in the Ascidian. 



A curious error is, it seems to me, made by those who 

 hold that the simplication or " degeneration " which we 

 see in the Ascidians and in a good many other animals is 

 due to a true retrogression in the structure of the 

 degenerating organism, the latest step in previous pro- 



