February 3, 1922] 



SCIENCE 



109 



failure of certain types of zygotes to develop 

 a succeeding generation. Explanations for all 

 of these conditions may be offered by postu- 

 lating lethal factors, as suggested by the work 

 on Drosophila, but it is well to understand 

 for plants how various are the ways in which 

 lethal factors maj' block the course of develop- 

 ment and how numerous are the points at 

 which they may operate. 



The significance of the impure species and 

 the importance of its place in certain natural 

 groups is not j'et appreciated. Curiously the 

 plant most conspicuously brought to the front 

 as one giving rise to new species by mutation 

 has become one of the forms most thoroughly 

 studied as an example of an impure species. 

 I refer of course to the plant QSnothera 

 Lamarckiana. Presented by De Vries as the 

 best illustration of his view that pure species 

 at times pass through periods when they 

 actively produce by large saltations new spe- 

 cies, the status of (Enotliera Lamarckiana from 

 the first became a subject for sceptical exam- 

 ination on the part of a body of naturalists 

 who hesitated to accept De Vries' conclusions, 

 and sought for other hypotheses to account for 

 its remarkable behavior. Bateson was the first 

 to suggest that the fifty per cent, or more of 

 pollen sterility in Lamarckiana indicated a 

 hybrid constitution. Jeffries pushed this argu- 

 ment with force through comparisons of pollen 

 sterility in Lamarckiana with similar condi- 

 tions in various known hybrids. Workers 

 with OEnothera now generally recognize for 

 most of their material the presence of very 

 high degrees of sterility both gametic, as indi- 

 cated by bad pollen and abortive ovules, and 

 zygotic, as shown by large proportions of 

 seeds incapable of germination. Renner has 

 recently taken the subject of pollen analysis 

 to a new level by showing that genetic classes 

 of pollen may be distinguished in Lamarckiana 

 and in some other Oenotheras by differences in 

 the form of the starch grains within the pollen 

 cell and pollen tube. Cytological studies of 

 Gates, Lutz, Stomps, Hance, van Overeem and 

 others have shown that certain of the variants 

 thrown by Lamarckiana differ from the parent 

 type in having higher chromosome numbers 

 due to non-disjunction. This non-disjunction 



seems correlated with a loose association of 

 chromosomes in Lamarckiana and other Oeno- 

 theras that favors irregularities of chromo- 

 some distribution at meiosis such as may be 

 expected in hybrid material. Much breeding 

 evidence, chiefly from the work of De Vries, 

 has made it clear that Lamarckiana and other 

 Oenotheras develop two or more classes of 

 fertile pollen grains which give in various 

 crosses sets of hj'brids in pairs, in threes and 

 in fours, good evidence of hybrid behavior. 

 I have shown that with care in the selection of 

 parent stock it is an easy matter to synthesize 

 a large-flowered vigorous hybrid with so many 

 points of resemblance to Lamarckiana that it 

 would be difficult to separate in descriptive 

 botany the hybrid from the assemblage of 

 biotypes that pass under the name Lamarckiana 

 which, as Heribert-Nilsson has so well brought 

 out, represents a collective species. Further- 

 more, this hybrid, an impure synthetic species, 

 which I have called neo-Lamarckiana, has 

 thrown in each of six generations from selfed 

 seed similar sets of marked variants, and, as 

 pollen parent in appropriate crosses, gives 

 twin hybrids thus paralleling in essentials the 

 characteristic performance of Lamarckiana. It 

 is of interest that among the variants from 

 neo-Lamarckiana there appear occasional 

 triploid and quadriploid forms comparable to 

 semi-gigas and gigas. There is no reason to 

 expect that neo-Lamarckiana will ever be 

 other than an impure species no matter how 

 close may be the inbreeding and selection to 

 type. It breeds true through only a small 

 proportion of its progeny and we can see 

 nothing that might change this habit so long 

 as the line lives. Finally, against the assump- 

 tion that Oenothera Lamarckiana is a pure 

 species is the fact that the plant is unknown 

 as a wild species and there is strong prob- 

 ability that it arose as a hybrid in England 

 about the middle of the last century. 



These are some of the reasons why genet- 

 icists rather generally have come to the con- 

 clusion that (Enothera Lamarckiana is repre- 

 sentative of an impure species which repro- 

 duces its heterozygous constitution because the 

 viable zygotes produced are for the most part 

 only those resulting from the union of two 



