30 THE EVOLUTION OF THE METAZOA 



SO far ("a living fossil"), or we succeed in establishing new 

 facts— facts which had not been known previously— through 

 our study of the structure and of the ontogeny of a known 

 animal species. The new fact must throw a new light on our 

 present concepts. The second way that can be taken which, 

 even if harder, is in reality much less dependent on a mere 

 chance, is that of a profounder comparison or of a juxtapo- 

 sition of facts that are already known. It enables us to discover 

 new relationships. Frequently we find this second w^ay to be 

 connected with the discovery of mistakes made by earlier 

 authors when they had formulated their final conclusions or 

 in relation to methods which they had incorrectly used. This 

 second way has been pursued in our investigations. 



In our case, that is in the study of the origin of Cnidaria, 

 it is impossible to expect that we will be able either 

 to find the corresponding fossil-material or to make 

 direct observations; and so it is clear that we must rely on 

 the comparative method as applied to the present living 

 species. The fossil documents that have been found belong 

 exclusively to the skeleton-building Cnidaria: with a few 

 exceptions they are all impressions of fully-developed 

 medusae. We can see, however, that even in those cases 

 where a much richer fossil-material stands at our disposal 

 (e.g. in Vertebrata) and where we are able, on the basis of 

 the fossil-remains of skeletons, to make far-reaching conclusi- 

 ons with regard to the structure of the "soft" parts of their 

 bodies, that even here we must use the comparative method 

 in its full extent in order to be able to make correct interpreta- 

 tions of facts i.e. to reconstruct the way the evolution had 

 taken. 



It should be mentioned in passing that only humble results 

 have been achieved so far by means of the biochemical method 

 in our study of the degrees of relationship, even if considerable 

 expectations had been originally placed in this method. The 

 chemical processes in question are certainly very complicated 

 and the special serum methods that have been used till now 



