ON THE ORIGIN OF SPECIES. 495 



176 specimens. The whites were very weak, and all of them died with- 

 out flowering. But they reappeared each year, and in 1897 I suc- 

 ceeded in getting them to blossom. After that they proved to be con- 

 stant and true to seed. The same is true for the red-nerved ones and 

 for the oblong as, which are very typical and easily cultivated species. 



Since 1895 I have each year sown Oenothera Lamarchiana, always 

 taking care that the seed was pure. Fertilization was always artificial, 

 with their own pollen, and with the exclusion of all insects. Yearly 

 I had a thousand or more seedlings and regularly found among them 

 a number of mutations. The new forms with which I already was 

 acquainted reappeared each time; with a single exception no others 

 have been added. The percentage of mutants remained the same each 

 year, of course with slight variations. 



Eepeatedly I saw new species originate which either did not flower 

 or were sterile or which on account of general weakness succumbed 

 early in life. Some of these clearly originated several times, others so 

 rarely that it was practically impossible to make a diagnosis. A few 

 of these I also found in the original locality. Hence nature evidently 

 makes besides species capable of existence also those which are not so. 

 The latter disappear very soon and hence are hardly ever seen; the 

 former persist for a greater or smaller number of years. 



The above may be considered sufficient to prove that the origin of 

 species is a phenomenon falling entirely within the limits of ordinary 

 observation. One has but to search his surroundings for a plant which 

 happens to be passing through a mutation period to be able to study 

 the entire process. Transportation to the garden only serves to make 

 isolation of the plant possible; it but shows what happens in nature, 

 but which there, on account of unfavorable conditions, is but seldom 

 or imperfectly observed. 



At the same time one sees that experiment, in this first example, 

 confirms the deductions made a long time since from paleontological 

 and biological data. 



Delboeuf, as well as Scott, requires that each new species does not 

 appear in a single specimen, but in a number of specimens, and not 

 once but during a number of years. For only under these conditions 

 are their chances sufficient. It is exactly this which happens with the 

 Oenotheras. They are formed each year, 1 per 1,000 or 1 per 100, in 

 any case in a sufficiently large number to fall within the requirements 

 formulated by the savants just mentioned. They are with a single 

 exception at once constant from seed, without ever returning to the 

 type of the mother species; they would, by sufficient isolation, at once 

 form groups of uniform individuals. Nothing indicates their appear- 

 ance in advance, there is not even a hint of transition; once formed 



