C— GEOLOGY 73 



However complete and specific the recapitulatory record may be at 

 the outset, in subsequent generations it becomes curtailed as the result 

 of increasing acceleration in individual development. This leads to the 

 ' skipping of stages ' either by a ' mere shortening of ontogeny,' in cases 

 where evolutionary trends remain constant, or by a ' straightening of 

 ontogeny ' where new trends out of accord with the foregoing set in. 

 The record also becomes vitiated as the result of the fact that acceleration 

 is not constant for all features, and consequently the combination of 

 characters exhibited by the adult ancestor becomes broken up, or even 

 eliminated from the development of the descendants. 



Turning now to the juvenile features which characterise the embryonic 

 stages of development, we find the problem is more difficult of elucidation. 

 That these features, exhibited during the early development of the 

 ancestor, are repeated during the corresponding or slightly earlier stages 

 of the descendants may be regarded as established. The unsettled 

 problem is the extent to which they reflect characters which in primeval 

 times were in the first place peculiar to the adult, or had their first onset 

 only in the embryo. 



Prospective Aspect. 



The task that lies before us now is to inquire the ways in which 

 evolutionary changes may be foreshadowed during development. For 

 this purpose our attention must be concentrated upon the new features 

 which mark those changes, upon the mode and time of their appearance, 

 and upon the way in which they fit into that framework of anciently 

 derived characters discussed above. 



In dealing with this aspect there is no need to stress once more the 

 importance of basing conclusions upon ample evidence made up of 

 numerous specimens precisely dated. It should, however, be urged that 

 specimens used for developmental studies must be selected from among 

 those which lie upon or close to the mode of the frequency distribution 

 curve for varieties occurring at each stratigraphical horizon. If this be 

 not done, conclusions of an extraordinarily contradictory character may 

 be drawn. This may be illustrated by reference once more to the gens 

 Zaphrentis delanoiiei. As already seen, specimens selected from the 

 mode for varieties collected from the Cementstone and Lower Limestone 

 horizons belong to Z. delanouei and Z. constricta respectively, and the 

 development of the latter faithfully recapitulates the adult condition of 

 the former. On the other hand, if by any chance the specimens selected 

 happened to be varieties at opposite extremities of the curve, viz. 

 Z. constricta from the lower horizon and Z. delanouei from the upper 

 horizon, then the young stages of Z. constricta would appear to anticipate 

 the adult condition of Z. delanouei and therefore to support the principle 

 of proterogenesis, which will be discussed below. 



In reading Garstang's discussions of kindred problems from the 

 biological side, I find myself very largely in agreement with him as far as 

 the evidence at his disposal takes him. Any difference that exists between 

 us seems to me to be due to the fact that he had not before him any 

 satisfactory evidence for the recapitulation of adult stages. He suggests 



D 2 



