THE GEN"EALOGY OF SHIPS. 8l3 



the conditions remain essentially unchanged, why should 

 a new origination he expected to have a structural con- 

 formation fundamentally different from its predecessors? 

 The structure must remain fundamentally similar, even 

 if it he a new origination; and thus, as Agassiz used to 

 argue, a relationship of thought would be seen to run 

 through the whole series, binding it in a unity as real 

 as if threaded on a genealogical line. If, then, a series 

 of independent originations 'is something so possible, and 

 so defensible, is it not plainly a petitio j^f^iiicipii to assume 

 that genetic relationship is the only possible explanation 

 of the graduated successions revealed in the history of the 

 equine type? 



There was another ground of hesitation to accept ge- 

 netic relation as the necessary explanation of equine suc- 

 cession. So far as conclusive evidence had gone, genetic 

 relationship was one of approximate identity. Inheritance 

 meant reproduction or continuance of the same specific 

 type. True, indeed, tliat susceptibility of variation co- 

 existed; but all observed variation among individuals of 

 the same ancestry exhibited but narrow structural range, 

 and seemed to tend to disappearance. But the range of 

 structure from Oyohippus to Eqiius was great. They 

 were not only different species, but different genera. Who 

 had ever known generation to wander so far away from 

 a primitive pattern? There existed, in truth, a real tvant 

 of analogy between the relationships in the terms of the 

 equine series and those in the generations of an estab- 

 lished lineage. 



One point only was suggested by the respondents which 

 tended positively to turn attention toward genetic descent 

 as the true explanation of equine relationships. " M. B. B." 



