28 INTRODUCTION. 



while the blastula stage, Fig. 2, VIII, reached at the close of 

 segmentation, bears a striking resemblance to such adult forme 

 as Volvox or Pandorina. 



There is, however, another side of the question which must 

 not be overlooked. Although it is undoubtedly true that deve- 

 lopment is to be regarded as a recapitulation of ancestral phases, 

 and that the embryonic history of an animal presents to us a 

 record of the race history, yet it is also an undoubted fact, 

 recognised by all writers on embryology, that the record so 

 obtained is neither a complete nor a straightforward one. 



It is indeed a history, but a history of which entire chapters 

 are lost, while in those that remain many pages are misplaced, 

 and others are so blurred as to be illegible ; words, sentences, 

 or entire paragraphs are omitted, and, worse still, alterations or 

 spurious additions of later date have been freely introduced, and 

 at times so cunningly as to defy detection. 



Very slight consideration will show that development 

 cannot in all cases be strictly a recapitulation of ancestral 

 stages. It is well known that c losely all ied_.&nimals_max differ 

 markedly in their modes of development, which could not be 

 the case if both recapitulated correctly. The common frog, for 

 example, is at first a tadpole breathing by gills, a stage which 

 is entirely omitted by the little West Indian fr_o_g, Hy lodes. A 

 crayfish, a lobster, and a prawn are allied animals, yet they 

 leave the egg in totally different forms. Some developmental 

 stages, as the pupa condition of insects, or the stage in the 

 development of a tadpole in which the oesophagus is imperforate, 

 cannot possibly be ancestral. Or again, a chick embryo, of 

 say the third day, Fig. 113, is clearly not an animal capable of 

 independent existence, and cannot therefore correctly represent 

 any ancestral condition ; an objection which applies to the 

 earlier developmental histories of many, perhaps of most, animals. 



HaBckel long ago urged the necessity of distinguishing, in 

 actual development, between those characters which are really 

 historical and inherited, and those which are acquired or 

 spurious additions to the record. The former he terms palin- 

 genetic or ancestral characters, the latter cenogenetic or acquired. 

 The distinction is certainly a true one, but an exceedingly 

 difficult one to draw in practice. The causes which prevent 

 development from being a strict recapitulation of ancestral 



