20 HiCKLING, Variation of Planorbis niultiformis, Bronn. 



earlier than that at which the type represented makes its 

 first appearance. Incidentally, there should be noted the 

 importance of studying the proportions in which the several 

 variants of a species are present in a given stratum. 



In conclusion, I would point to the striking testimony 

 borne by the history of these shells to the production of 

 extensive morphological changes by a process of purely 

 continuous variation. In this case it is clearly quite im- 

 possible to introduce any idea of " mutation," at least in 

 the de Vriesian sense, into the process. No one can 

 indicate the point at which a new " form " or " type " or 

 "mutant" appears, the sequence being quite unbroken. 

 And surely the great mass, if not the whole weight, of 

 palaeontological evidence points in the same direction ? 

 Nothing, it appears to me, could be more dangerous, or 

 more subversive of the logical basis of stratigraphical palae- 

 ontology, than the introduction of the idea of mutation. 

 If species or varieties arose in that manner we must 

 assume that the date of their appearance can be defined 

 with absolute precision, since there can be no fine stages 

 of modification leading gradually up to them ; if the con- 

 trary is true, there can be no absolute palreontological 

 boundary lines. It must be long before the relative 

 importance of these two factors in evolution can be finally 

 settled, but in the meantime one case in which continuous 

 modification is clearly demonstrable as a palaeontological 

 fact should outweigh many instances in which new t\-pes 

 appear without ascertained intermediates linking them to 

 their parent species. 



