NA TURE 



[November 26, 1908 



them have come, and partlj' because, even if this 

 were known, it would still not be known whether 

 thev were the result of a self- or cross-fertilisation. 



CEnotheras were therefore transported from the 

 field at Hilversum to the garden at Amsterdam. 

 This was done in one of two ways. Either the young 

 first year's rosettes were transplanted (CE. 

 Lamayckiana generally behaves as a biennial, flower- 

 ing in the year after that in which it was sown), or 

 seed was collected from the mature plants in the 

 field at Hilversum and sown in the garden at 

 Amsterdam. 



The expectation, based on the appearance of two 

 new species in the field, that more would arise in the 

 garden was fulfilled. There arose altogether about 

 a dozen new elementary species in the garden. The 

 vork of investigating the mode of origin of these 



tlon of CEnotheras from Hilversum to Amsterdam, 

 was to sow the seed directly in the bed in which the 

 plants were to flower. The disadvantage of this 

 plan was that all the seed did not come up in the 

 first year ; so that, in the first place, all the crop 

 arising from a single sowing could not be recorded 

 at one time, and, in the second, the bed could not be 

 used for another sowing until it was certain that all 

 the seed from the last had come up. Moreover, 

 weedinc and the minute examination of the seedlings 

 was not by any means convenient in these circum- 

 stances. This plan was therefore soon forsaken, and 

 that of sowing the seed in pans adopted in its stead. 

 The pans were filled with soil which had been baked, 

 a process which killed any seeds which might be in 

 the soil, so that there was no possibility that any of 

 the CEnotheras which came up could have arisen 



new species consisted partly in finding out if the 

 relative numbers of these species appearing every 

 year were at all constant, and, if so, what the " muta- 

 tion coefficient" (as this number was called) was; it 

 consisted also in testing the constancy of each new 

 species through several generations. Besides this, 

 de Vries was continually on the look out for new 

 species, and for this purpose large sowings of 

 CEnothera seed were made every year. Moreover, 

 crossings between the various elementary species 

 were continually being carried out. The number of 

 plants which had to be examined in the course of this 

 work was enormous ; and the number could not have 

 been so great, nor the work so thorough, if de Vries 

 had not paid special attention to the distinguishing 

 characters of the seedlings of the various species. 

 His plan at first, i.e. shortly after the transporta- 



NO. 2039, VOL. 79] 



his gtaenho 



from any other source but the seeds deliberately sowi; 

 in the soil, and the tiresome and difficult process of 

 weeding was rendered unnecessary. The seedlings 

 could be examined much more minutely and 

 thoroughly in the pans than in the beds. 



The result of this innovation was that de Vries 

 acquired a most intimate familiarity with the seed- 

 ling characters of . the various new elementary 

 species ; this is to a certain extent putting the carr 

 before the horse. It is perhaps truer to say that the 

 majority of the new species which were discovered 

 after the introduction of this innovation owed their 

 discovery to the fact that they differed from the 

 parent form in the seedling stage. 



Perhaps the most valuable improvement in the 

 equipment of the Experimental Garden was one 

 which was made possible by the practical way in 



