394 



METAMORPHOSIS 



Taking origin from a solitary example, 0. gigas forthwith appears with constant 

 and pure seed ; arising at a bound from the parent form, in a moment a 

 new species sprung into independent existence. In this way also the other 

 species I have described arose, suddenly and without any transitional forms.' 



There are a whole series of questions which are closely connected with this 

 observation of De Vries, and of these we can touch on only a few of the more 

 important. First of all, how can a sub-species arising from a single specimen main- 

 tain itself in nature ? If we assume that crossing with the type takes place, the 

 species would soon disappear, unless its characters were absolutely predominant; 

 further, no subsequent segregation is possible, as in the case of the hybrid peas 

 quoted previously. That something of this kind may happen is shown by 

 the experiences of the American farmer with the Ancon sheep. There was born 



Fig. ii6. Oenothera lamarckiana. .\pex of 

 a shoot at the commencement of flowering ; b, 

 withered flower lying on the subtended leaf. After 

 De Vries (1901). 



Fig. 117. Oenothera gigas: \ derived from Q. la- 

 niarckiana in 18Q5. .'\pex of the shoot at the com- 

 mencement of flowering ; at a, a petal has been re- 

 moved ; b, withering flower. After De Vries (1901). 



in Massachusetts in 1891 a lamb with the form of a dachshund. In spite of 

 cross-breeding with the parental stock the Ancon type still predominated, and 

 the race increased without any selection at all, artificial or natural (Darwin, 

 1868). If we assume self-fertilization, a maintenance of the new species can 

 only take place if it be more dominant than the parent species, and gradually 

 drives it out. The gregarious occurrence of closely-related species would, under 

 these conditions, be difficult to explain. Such being the case, another possibility 

 must be admitted. De Vries observed, in the case of the new species of 

 Oenothera, that they arose not once but several times from the parent stock. 

 It seems probable that the old species produced the new ones in ever-increasing 

 numbers. We certainly cannot advance any proof of this assumption, but it 

 may be supposed that the originating of a mutation induced besides an internal 



