442 JSXPEBJMENTS WITH PLANTS 



in nature, just as forms originate in the garden, by 

 the selection of ordinary fluctuating variation alone. 



In other words, an ordinary fluctuating variation 

 may be so intensified by selection as to form a distinct 

 mark of difference between the improved and the 

 original form: when all the intermediate forms have 

 perished, we have two species instead of one. Darwin ^ 

 believed that species originated in this way both in 

 the garden and in nature (where the struggle for 

 existence selects the fittest and destroys the rest) , 

 and that the species so formed remain distinct. 

 Professor de Vries has recently called this view in 

 question on the basis of some very remarkable experi- 

 ments.^ 



Selection cannot make new species, he declares, 



1 Darwin likewise believed that species could origiuate by sudden varia- 

 tion, but was inclined to lay less emphasis on this mode of origin. 



- In these experiments ( which hare lasted over twenty years ) 

 Professor de Vries has comliined in brilliant fashion the special points of 

 superiority of various methods. The plant - breeder has the advantage 

 of being able to grow and .iudge large numbers of individuals in a limited 

 space: of this he has made use. The animal -breeder in the most important 

 cases keeps in a book a pedigree -record of eacli individual animal, its 

 characters and those of its offspring. Professor de Vries has also done 

 this in all the most important cases, giving a number to each bed, other 

 numbers to each row in the bed, and also to eacii plant in every row. 

 Thus, each plant receives ii number, and a record is kept both of it and 

 its offspring. In many cases the seeds of each separate plant were sown 

 in separate beds, in order that the hereditary percentage might be clearly 

 determined. Most careful precautions have been taken to prevent crossing, by 

 enclosing the flowers in parchment paper bags. Finally, he has made 

 extensive use of the statistical methods described above (p. 417) and thus 

 discovered many important principles which would not otherwise have 

 come to light. 



