I9I5.] PURE SPECIES OF (ENOTHERA. 227 



(De Vries '13, Davis '14). The zygotes of a pure species must 

 be uniform since the gametes of each sex are respectively similar, 

 and a pure species, to employ that convenient expression of 

 Bateson's, is therefore homozygous. 



It has generally been held that no further proof of the genetic 

 purity of a species is necessary than the established fact that it will 

 " breed true," and I venture to believe that at present most workers 

 among the Oenotheras regard this test as entirely sufficient to establish 

 the character of any material with which they work. If any line of 

 CEnothera breeds true in large cultures it is confidently regarded as 

 homozygous. Should a line fail to breed true to any considerable 

 degree it is stamped as a hybrid if the investigator inclines towards 

 the methods of analysis characteristic of the Mendelian school. 

 Those who believe in mutations are so fully content with this test 

 that to them a form need breed only reasonably true to pass as a 

 pure species and the departures from the type, called mutations, are 

 interpreted as due to modifications of the germ plasm not, however, 

 the result of hybridism. 



If a line of CEnothera fails to breed true to a very considerable 

 degree and thus becomes suspected of a hybrid constitution, few 

 workers would think of using it as favorable material for experi- 

 mental studies to test the mutation theory. It is the lines which 

 breed reasonably true that chiefly form the subjects of CEnothera 

 discussions with reference to the theory of mutation. Such a line 

 is the Lamarckiana of De Vries's cultures which when grown in 

 large numbers in selfed families appears uniform except for certain 

 small proportions of individuals, " mutants," which stand out clearly 

 from the mass with distinctive characters that are readily recog- 

 nized and may be clearly described. It is important to note that 

 these new types are not connected by intergrading forms with the 

 parent Lamarckiana and that they appear in successive generations 

 of Lamarckiana with certain degrees of regularity. 



More impressive than this history of Lamarckiana which has 

 flowers open-pollinated, and consequently likely in Nature to have 

 been crossed by insects, is the behavior reported for certain lines 

 of CEnothera with flowers close-pollinated in the bud, a condition 

 that obviously gives their own pollen the first chance to function and 



