Conclusion. 499 



not an intrinsic character of a particular species, but a 

 passing phase in which the plants in a particular locality 

 may happen to be. 



Only one species answered my purposes: Oenothera 

 Lamar ckiana. Even on the spot where I found it, it gave 

 promise of a more favorable result than all the rest. In 

 the first place, it was not a genuine wild form, but an 

 escaped one, which had spread from a bed to a deserted 

 field close by, where it had multiplied abundantly. A 

 rapid multiplication of this kind is one of the supposed 

 causes of mutability. In the second place it was very 

 rich in partial abnormalities ; not only in the common 

 ones like floral anomalies, pitcher formations, fascia- 

 tions, connations, adnations, etc., but in the rarer ones, 

 like the development of secondary axil buds in the in- 

 florescence and so on. In the third place I found iso- 

 lated delicate plants with narrow leaves which only 

 formed rosettes and then died. I was unable to studv 

 them further at the time but they have since turned o\\\ 

 to be a perfectly good new species (0. elliptica). And 

 lastly I found there two well characterized forms which 

 were hitherto unknown and have since proved to be con- 

 stant (O. lacvifolia and O. hrevistylis) . 



But the result of sowing the seed collected in the 

 field, was decisive. I did this first in 1887 and repeatedly 

 afterwards, but particularly in 1889. IMy very first cul- 

 ture gave me what I wanted ; it contained a form, which 

 differed sharply from the normal in almost every feature, 

 which had not been seen in the field before and was other- 

 wise also absolutely unknown. This was the Oenothera 

 lata. In the following year I sowed the seeds of plants 

 which I had brought with me from Hilversum in the 

 autumn of 1886 as rosettes: they gave me the same form 



