Mutations in Nature. 303 



a given species in a certain locality is in a mutable con- 

 dition is to collect its seeds and sow them. The sowing- 

 should be carried out on a large scale. A whole series 

 of experiments, which I have started with a number of 

 different species with this end in view, have been without 

 any positive result. From which I conclude that muta- 

 tions in nature are rare ; although I am convinced that 

 they will be found from time to time if they are carefully 

 looked for. 



I have for many years applied both these methods to 

 the case of Oenothera Lamarckiana. I have visited the 

 field itself almost every year, or, if not, have had it vis- 

 ited by others. The majority of the new forms have 

 been observed in that way, but usually as weak seedlings 

 or young rosettes; and very rarely indeed in flower. 

 Furthermore I have collected seed in the field, especially 

 in the period 1886-1888, when I began my experiments, 

 and have sown it in my experimental garden; in the 

 first two years in small quantities, but in the last one on 

 a large scale. Since then I have repeated the experiment 

 from time to time and only stopped when I was pretty 

 certain what would happen. 



I shall now give a list of the various species which 

 I either found in the field at Hilversum or raised from 

 seeds which I collected there. 



Oenothera lata. In 1887 I raised three examples of 

 lata which were all annual from seeds which I gathered 

 in the autumn of 1886 from quinquelocular fruits of 

 otherwise normal Lamarckiana plants. Two stayed in 

 my garden and gave rise to one of the /afa-families de- 

 scribed in § 7 (p. 288) ; the third germinated with three 

 cotyledons and was transplanted to a garden near Hil- 

 versum where it flowered but did not set fertile seed. 



