ix KECAPITULATION 493 



Again in considering whether a particular feature of structure 

 is to be regarded as ancestral or as a modern adaptation the following 

 questions should be asked : (1) Is the feature peculiar to one group 

 of Vertebrates or does it occur in several groups, and (2) if it occurs 

 in several groups do the various animals possessing the peculiarity 

 in question undergo their larval stages in similar sets of environ- 

 mental conditions ? 



If the particular feature occurs in several groups derived from 

 a common ancestral form this obviously increases the probability 

 of the feature itself being ancestral. If however the several groups 

 show similar sets of environmental conditions during their larval 

 stages this introduces the element of doubt whether the similar 

 features may not after all be merely adaptations to these similar 

 sets of conditions. 



Again it is important to make out whether the particular 

 similarity has to do merely with parts of structure in direct func- 

 tional relationship to external conditions. If there be deep-seated 

 correspondences in structure with no such direct functional 

 relationship to external factors then this gives greatly increased 

 probability to these correspondences being truly ancestral in their 

 nature. 



The morphologic in trying to decipher the record of evolutionary 

 history from the data of comparative anatomy or embryology is 

 constantly impressed by the potency of nature's economy of living 

 substance. An organ no longer required may be eliminated within 

 a very short period of evolutionary time. Thus in some species of 

 Mackerel {Scomber) so important an organ as the air-bladder has been 

 eliminated : in various Frogs and Toads the external gills have been 

 eliminated from development. Thus negative embryological evidence 

 is of peculiarly little weight in relation to phylogenetic problems. 



(5) The Protostoma Hypothesis. — This is a working hypothesis 

 which links together and in a sense explains a number of features in 

 the early development of Vertebrates which are otherwise extremely 

 puzzling. The more important of these features may be summarized 

 as follows : 



In AmpMoxus as has already been shown the dorsal side is at 

 first occupied by the widely open gastrular mouth. Later this 

 becomes roofed in by a backgrowth of the gastrular rim anteriorly. 

 A similar process of backgrowth appears to take place in the gastru- 

 lation of lower Vertebrates in general. The roof of the gastrular 

 cavity formed by this process gives rise later not merely to the 

 dorsal wall of the alimentary canal but also to notochord and central 

 nervous system. 



I. Now occasionally there are appearances which suggest that this 

 archenteric roof consists really of two lateral halves which have 

 become fused together along the sagittal plane. Thus in Protopterus 

 the down-o-rowing dorsal lip of the blastopore is frequently indented 

 by a median incision, Again in Urodeles the medullary plate is 



