i893. BIOLOGICAL THEORIES. 199 



The statement that this stage in the development of a bird or a 

 mammal is the modified remnant of the ndnlt structure of the ancestor ; 

 and that the ontogeny is even in part made up of a series of remnants of 

 ancestral adiilt structures arranged in chronological order, is not only 

 unjustified, but is demonstrably false. If any generalisation in the 

 whole science of zoology is borne out by fact, it is the law of Von 

 Baer with reference to animals living free only in later stages of 

 development. That law claims a parallelism between the develop- 

 ment of a fish and of a bird which is quite inconsistent with the re- 

 capitulation theory, and completely consistent with the observed facts. 



The early stages of the fish embryo are very like those of the 

 bird embryo. These two do correspond to each other. The state- 

 ment that the embryonic structure of a bird follows a course 

 which is, from beginning to end, roughly parallel with, but somewhat 

 divergent from, the course followed by a fish, is borne out by the 

 actual facts. A bird does not develop into a fish and then into a 

 reptile and then into a bird. There is no fish stage, no reptile stage, 

 in its ontogeny. The adult resembles an adult fish only very remotely. 

 Every earlier stage resembles the corresponding earlier stage of the fish 

 more closely. There is a parallelism between the two ontogenies. 

 There is no parallelism between the ontogeny and the phylogeny of 

 either a bird or any other animal whatever. A seeming parallelism 

 will fall through when closely examined. 



Each transient stage in the development of any individual is a 

 modification of the corresponding stage of development of its ancestors. 

 It is in no case a modification of the adult stage of the ancestor. The 

 adult stage of a bird, and no other, corresponds to the adult stage of 

 the fish-like ancestors (if it ever had such ancestors). 



The stalked "pentacrinoid" larva of Antedon (^ = Coniatnla) is 

 a modified equivalent of the stalked larva of the " pentacrinoid " 

 ancestor (if ever there was such an ancestor), and not the modified 

 equivalent of the adult ancestor. The possession of a stalk in early 

 stages of development appears to be an advantage, and hence the 

 specific constitution which determines the development through a 

 stalked stage has been preserved by Natural Selection. There is no 

 evidence whatever to justify such mystical conceptions as those 

 involved in even the most reasonable forms in which the recapitulation 

 theory has been applied to this case. 



The promise which the theory gave of serving as the guide to 

 knowledge of past history, without the labour involved in palaeon- 

 tological research, was, indeed, tempting : and when the " royal road 

 to learning" had been shown by it, it is not surprising that some 

 zoologists should have entered for the race along this road. To what 

 goal that road has led may be learned by a comparison of the 

 numerous theories as to the ancestry of " chordata " which have been 

 put forward by those who adopted the theory without enquiring as to 

 its validity. 



