228 DAVIS— THE TEST OF A [April 23, 



thus greatly reduces the probabilities of cross-pollination. Such a 

 plant is the biennis of Holland and other parts of Europe, a type of 

 especial interest not only for its clear morphological characters but 

 also because there is good reason for believing the line to be very 

 old. This plant forms a large population in Holland with no near 

 relatives and must have lived there for many years to have so thor- 

 oughly established itself. Indeed it seems probable that this 

 Oenothera, the Dutch biennis, has come down to us essentially un- 

 changed from the times of Linnaeus who gave us its name. We 

 know of no plant better representative of a species of Oenothera 

 and we know of no CEnothera which better satisfies the generally ac- 

 cepted requirement that a species should "breed true." 



CEnothera biennis L. in large cultures comes so true that hun- 

 dreds of plants may be grown without finding a single departure 

 from the type. Yet Stomps ('14) in large cultures of selfed lines 

 from a single wild plant collected in 1905 discovered that this Dutch 

 biennis throws occasional marked variants (" mutants ") and he de- 

 scribed a biennis semi-gig as with the triploid number of chromo- 

 somes (21), a dwarf type biennis nanella, and a color variety 

 biennis sulfurea with pale yellow petals. De Vries ('15) at once 

 took up the study of certain of the lines established by Stomps and 

 grew cultures which totaled 8,500 plants. Among these were 4 

 plants of biennis semi-gigas about 0.05 per cent., 8 plants of biennis 

 nanella about o.i per cent., and 27 plants of biennis sulfurea about 

 0.3 per cent. Since the percentages from Lamarckiana are for 

 semi-gigas 0.3 per cent, and for nanella i to 2 per cent, it should be 

 noted that with respect to these " mutants " biennis appears to be 

 the more stable of the two species, although the color variety biennis 

 sulfurea constitutes a new type of variant in experimental studies 

 on Oenotheras. A culture of over 1,000 plants from selfed seed of 

 biennis sulfurea, all with pale yellow flowers, produced 2 dwarfs 

 thus establishing a "double mutant" O. biennis mut. sulfurea mut. 

 nanella. 



As evidence for the mutation theory of De Vries this behavior 

 of the Dutch biennis is to the writer much more trustworthy evi- 

 dence than the behavior of Lamarckiana for the reason that the 

 latter plant in his opinion does not have a clear record of long 



