726 



SCIENCE. 



[N. S. Vol. XV. No. 384. 



this genus, I might combine them in the A species therefore is not born only a 



form of a 



single time, but repeatedly, in a large num- 



Genealogical Tree of Oenothera Lamarckiana. 



0. Lamarckiana forms the main stem ; all 

 other species originated from its seeds. 

 Descendants of the mutations are not in- 

 cluded in the scheme, so as not to make 

 it too intricate. 



The first two generations shoA¥ed but 

 comparatively few types. The reason for 

 this may be sought in the fact that at the 

 time I did not know how to trace them. 

 Hence the fourth generation shows a 

 majrked improvement, which continued 

 after the sowing had undergone a great 

 numerical reduction. 



0. oblonga appeared by hundreds. All 

 of these plants closely resembled each 

 other. They could be recognized as 

 rosettes of root leaves by the narrow leaves 

 with broad veins, and later on by their deli- 

 cate, stiff, nearly unbranched, seemingly 

 naked stems. The same is true for the 

 dwarf forms. Our genealogical tree shows 

 of these about 150; in other experiments I 

 have met with even larger numbers. These 

 plants again form a distinct type, which 

 could readily be recognized, whatever the 

 age of the specimens. 0. rubrinervis, 0. 

 albida and 0. scintillans were far rarer, 

 but as a rule appeared each year, always 

 bearing exactly the same character. 



ber of individuals and during a series of 

 consecutive years. 



It is clear that this fact, so apparent in 

 my experiments, must be of enormous im- 

 portance in the case of wild-growing plants. 

 'How small is the chance of a single plant 

 to triumph in the struggle for existence! 

 Only when a number, or rather a large 

 number, of similar individuals do battle 

 together for the same cause is it that this- 

 chance acquires a value. 0. gig as would 

 have been nipped in the bud were it not 

 for my aid. I have never found it grow- 

 ing wild, as I did some specimens of the- 

 less rare 0. lata, and 0. nanella. But these 

 also meet with too many hardships. Only 

 once have I found a single specimen in 

 flower. 



But next to the question of the more or 

 less frequent appearance of a new species 

 stands another which has as potent an in- 

 fluence upon its life. It is of course a 

 matter of pure chance whether a mutation 

 is or is not better adapted to the en-\dron- 

 ment than the parent species. Sometimes- 

 it will go one way, sometimes the other, 

 or both may be equally well adapted. Our 

 0. gigas and 0. rubrinervis are, during the 

 flowering period, as robust as the mother- 



