48 BIOLOGICAL LECTURES. 



ancestral inheritance, opened the door to an unduly subjective 

 way of dealing with embryological evidence and deprived the 

 method of that authoritative character which had so generally 

 been ascribed to it. Now the whole recapitulation theory is 

 boldly called in question, and, in the admirable lecture delivered 

 last year in this place, Prof. E. B. Wilson showed the untrust- 

 worthy nature of the embryological criterion of homology. The 

 difficulty in this case lies in the absence of any "canons of 

 interpretation " (to use Bateson's phrase) by which the con- 

 tradictory data of embryology may be harmonized into a con- 

 sistent whole. To take a concrete illustration : The ontogenetic 

 development of the horse's teeth would give us a very inade- 

 quate and indeed false conception of the actual steps of change 

 by which the modern type of dentition has been attained, nor 

 would embryology show that the horse is descended from five- 

 toed ancestors. Knowing, as we do from the fossils, the 

 phyletic series, the embryological facts may be readily under- 

 stood. It is an undue reliance upon such facts which has led 

 to the concrescence theory of tooth development now so rife 

 in Germany, and which seems so absurd when viewed in the 

 light of palaeontology. 



I have no intention of belittling the splendid services which 

 embryology has rendered to morphology, but merely to point 

 out that this method alone cannot reach finality any better 

 than comparative anatomy. It resembles dealing with a lit- 

 erature that has been vitiated by many forgeries, only the 

 grossest and most palpable of which can be readily detected. 



A third method of attacking morphological problems is that 

 offered by palaeontology. Let us begin our consideration of 

 this method by frankly acknowledging its drawbacks and limita- 

 tions, (i) In the first place there is the imperfection of the 

 geological record. Palaeontology does not profess and never 

 can hope to reconstruct the whole history of life upon the 

 earth, or even the greater part of that history ; very many 

 chapters are irretrievably lost, and others are so fragmentary 

 that they teach us little or nothing. The great sedimentary 

 deposits which contain nearly the whole recorded history of 

 the globe were laid down under water, and for a land animal 



