83 



illustrate the frequence and manner of occurrence of the mutants 

 derived from pure seed cultures of Onagra biennis {Oenothera 

 Lamarckiana) in a series extending from 1886 to 1899. 



In the genealogical table shown above, seeds from the nine 

 specimens of the first generation produced 15,000 of the parent 

 type, 5 of nanella and 5 of lata. Seeds from some of the 15,000 

 produced a crop consisting of 10,000 biennis, 3 nanella, 3 lata 

 and 1 rubrinervis. The succeeding generations were obtained in 

 the same manner. 



It is to be seen from the above table that in the series of cul- 

 tures outlined above, embracing seven generations of seedlings, 

 about 800 of the 50,000, or a little more than [.$fo of the entire 

 number, were mutants or forms sufficiently divergent from the 

 normal to be designated as new species. The parent type pro- 

 duced some of the new species every year it was under observa- 

 tion but by no means in the same proportion or profuseness, 

 and it seems very probable that no plant Avill exhibit the tend- 

 ency to produce mutants in greater degree than the one which 

 has been selected for these notable experiments. It is also to 

 be noted that the new species have by no means the strength and 

 general virility of the parent type, and that the few individuals 

 representing some of the new species in any community would 

 have but little chance of survival in the struggle for existence 

 with the thousands of their fellows of the parent type. When 

 isolated, however, and relieved from the fiercer competition met 

 under natural conditions, the majority were independent constant 

 types. 0. scintillans, O. subline aris and 0. elliptica were classed 

 by de Vries as inconstant forms, while 0. lata is sterile so far as 

 the examples yet examined show. 



A discussion of the facts given above could hardly be made 

 without calling up the question at once as to the systematic value 

 of the forms designated as species. The new species which sud- 

 denly originate do not differ so widely as an apple from a pear, 

 or as a pine from a spruce ; only in a few of the species are their 

 general features strikingly divergent from the parent type. Yet 

 a careful examination will show that differences are present 

 and important, relating to size and aspect of the shoot, shape, 



