208 LECTURES ON BIOLOGY 



species, 0. gigas. It is absolutely certain that its ancestors, at 

 least in the last three generations, were the ordinary 0. lamarck- 

 iana. The self-fertilized seeds of this TJrpflanze were in the 

 following year sown separately and produced more than 450 

 plants, which all exhibited the pure type of 0. gigas, there being 

 not one reversion to lamarckiana. Even in later generations 

 0. gigas remained constant. It seems therefore proved that a 

 new elementary species can appear suddenly in a single indi- 

 vidual and be constant from the beginning. 



While 0. gigas remained constant in its characters, there 

 were other mutations in which this fact was not observed. Such 

 an instance is 0. scintillans, from the seed of which, in spite of 

 self-fertilization, originated always three forms, 0. scintillans, 

 0. oblonga, and finally the original 0. lamarckiana. In addition 

 the seed frequently produced elementary species of a second 

 order which united in themselves the characters of two species. 

 Thus an intermediate form, for instance, between 0. scintillans 

 and 0. nanella, has not unfrequently been observed. 



Concerning the causes which effect the mutation, we can only 

 make guesses. We gain, however, the impression from the many 

 facts before us that mutability occurs periodically and is replaced 

 by periods of apparent constancy. As a rule the new ' elementary 

 species ' appear at once in a large number of individuals, an excep- 

 tion being formed by 0. gigas, of which only one individual was 

 observed. Like all organisms, 0. lamarckiana and its mutations 

 are subject to fluctuating variability, and one is able to produce 

 among them new races by rational selection. But they always 

 remain forms which are in ' need of selection ' ; they differ very 

 little from the type, probably only in one essential, and must not 

 be mistaken for the ' elementary species.' Of mutations, too, it 

 is therefore true that they can be either useful, or indifferent, 

 or injurious. Numerous mutations, like the sterile 0. lata, the 

 weakly nanella and albida, and the fragile rubrinervis, would 

 hardly be able to maintain themselves in nature. On the other 

 hand, 0. Icevifolia appears to be at least equal to 0. lamarckiana, 

 whilst 0. gigas certainly excels it. From these experiments 

 De Vries concludes that ' the mutations have no tendency ; some 

 of the new types perish without progeny. Among the rest of 

 fche newly formed and fully developed species natural selection 

 must later make its decision.' 



