THE MEANING OF EVOLUTION 257 



animal world, yet ^vithout this proof it could not be 

 contended that the veiled gastrula of the developing 

 frog's egg, for instance, is related genetically to the 

 gastrula of the Echinoderm larva. What experimental 

 embryology does indicate is that the formation of 

 gastrula and (in most groups) the three germinal 

 layers are only the means of morphogenesis. In the 

 division of the ovum, and the arrangement of the cells 

 to form the organ-rudiments, the formation of the 

 gastrula and the mesoderm are in general the line of 

 least resistance in the process of development. If they 

 do not appear, or are difficult to recognise in the 

 ontogeny of a group of animals, it is not a sound method 

 to assume their presence in an abbreviated or distorted 

 form, postulating that they ottght to be present, having 

 been transmitted by heredity. Physical conditions 

 undoubtedly influence developmental processes and 

 there is no reason for assuming that all ontogenetic 

 processes were originally the same. 



If we do not strain the facts of our descriptions of 

 organic nature, and if we do not build on unprovable 

 conjectures, all that morphology certainly shows us is 

 that the evolutionary process has led to the establish- 

 ment of some dozen or so great groups of organisms, 

 each with appended smaller groups more or less closely 

 related to them. Whether these greater lines of 

 descent are to be represented, as they usually are, as 

 branches springing from a single stem, or whether they 

 are truly collateral, each evolved independently of all 

 the others, is a question which is not to be solved merely 

 by the methods of comparative anatomy or embryology. 

 The widely different, and equally probable, phylogenies 

 of the past indicate that data for the solution of such a 

 problem do not exist, not just yet at all events. What 

 we may discuss with greater advantage is the question 



R 



