BIOLOGICAL PALAEONTOLOGY^ 91 



imperfection of the palaeontological record, phyletic or 

 morphogenetic conceptions inconsistent with the known 

 order of appearance of the phenomena concerned must 

 prove ill-founded, and cannot carry conviction. 



Valuable though this evidence is, two complicating 

 factors conspire to obscure the views of evolution 

 afforded. In addition to obvious gaps in the record 

 where periods, long or short, are locally unrepresented 

 by sediments, the recurrent differences in lithology of 

 most continuous stratigraphical series indicate physio- 

 graphical changes inimical to continuity of life. A 

 stock may have originated on sandy beaches in a 

 district where those deposits are no longer accessible ; 

 and it may have attained full differentiation before the 

 period of which strata, containing its remains, have been 

 preserved. So that in a varied rock-sequence, scraps of 

 the evolutional history of types inhabiting the several 

 environments are alone available. Even later discovery, 

 in distant regions, of suitable matrices accumulated at 

 the time of inception of the stock, cannot be relied upon 

 to supply the evidence needed ; since most, if not all, 

 new groups originate in restricted areas, extending their 

 geographical range only in later stages of progress. It 

 follows, therefore, that the beginnings, and to a less 

 degree the last phases, of race-history must usually be 

 lost; while continuous record of its acmaic episodes can 

 be pieced together only from dislocated fragments of 

 evidence obscured by local variation. In rock-systems 

 representing long-continued sedimentation under rela- 

 tively uniform conditions (such as the Chalk), fairly 

 extended and continuous views of evolutional sequence 

 are possible; it is from study of records preserved 

 under such circumstances that criteria can be deduced 

 for correlating the disconnected fragments of evidence 

 commonly available. 



