The Hypothesis of a Premutation Period. 495 



The genera Rosa, Rubus, Hieraciiim, Salix and some 

 other types rich in species formed the nebulous groui)s 

 of the older systematists : they were extremely rich in 

 forms that could hardly be distinguished. Without cul- 

 tivation we can only have a provisional knowledge of 

 their species and the cultivation of some of these forms 

 through even a few generations would be no light task. 

 But the richness of forms (exclusive of hybrids), is 

 comparable to that of Draba verna and Oenothera La- 

 marckiana, and clearly points to the existence of muta- 

 tion periods, belonging partly to the past and partly per- 

 haps to the present. 



The most conclusive point of all is however the ne- 

 cessity of assuming such a period to have occurred in the 

 group of Oenothera biennis (the subgenus Onagra), 

 which is absolutely analogous to the Laniarckiana-group 

 (see p. 440). 



Finally, a few words must be devoted to the ques- 

 tion as to when the premutation period occurred in our 

 particular case of the Oenothera Lamarekiana at Hilver- 

 sum. Two possibilities present themselves. Either the 

 plant was already in a period of mutation when its seeds 

 were first sown there by Mr. Six (in about 1870, see p. 

 266). Or the period began on the spot. In the former 

 case the Oenothera must have already been mutable and 

 this property had simply escaped observation. This 

 seems hardly likely because O. nanella, 0. gigas and O. 

 laevifolia, if they had come under the eyes of nursery- 

 men or amateurs, would certainly have been thought 

 worth cultivation, and would have been put on the market. 

 But no 'Varieties" of 0. Lamarekiana occur either in 

 descriptive works, or in catalogues. 



In the latter case; we should suppose that the rapid 



