46 BIOLOGICAL LECTURES. 



much weight is to be allowed a given similarity, and how far 

 this is offset by a dissimilarity which accompanies it, we have, 

 as yet, few means of determining, and have still to discover 

 those laws of organic change which shall render the same 

 service to morphology as Grimm's law has done to the study of 

 the Aryan tongues." 



Philology was raised to the dignity of a true science by the 

 laborious tracing back of modern words, step by step, to their 

 ancient origins through all their intermediate gradations, and 

 sound principles of etymology could not be established until 

 this was done. Morphology must profit by this lesson and 

 must imitate the method of the science of language. Not 

 until many long phylogenetic series have been recovered can 

 the law of change be worked out. It is just here that palae- 

 ontology is fitted to render invaluable services to the common 

 cause. 



As every one is aware, the principal methods of morphological 

 inquiry are comparative anatomy, embryology, and palaeon- 

 tology, each of which has its great advantages, but accom- 

 panied by its own peculiar drawbacks and limitations. Lack 

 of time will prevent any discussion of Bateson's proposed new 

 method from the study of variation. I have elsewhere exam- 

 ined that at some length. 



The foundation and corner-stone of the whole structure of 

 morphology must ever be comparative anatomy, an accurate 

 knowledge of which is indispensable to successful prosecution 

 of the other departments of inquiry. This method has, in the 

 hands of the masters, registered many great triumphs in the 

 solution of difBcult problems of homology, and of the mutual 

 relationships of animal groups. At the present time, the ten- 

 dency is to give more and more weight to its determinations. 

 On the other hand, finality cannot be reached by this method. 

 It suffers from the very significant drawback of possessing no 

 sure criterion by which to distinguish between those similarities 

 of structure which result from actual genetic relationship and 

 those which are due to parallel or convergent development, and 

 thus to determine the taxonomic value of a given likeness or 

 unlikeness. It is an exceedingly common fallacy to assume 



