THE EVIDENCE OF EVOLUTION. 371 



to overcome this difficulty. The theory of descent had to be remolded. 

 On this point conviction has grown in America during the last decades 

 with increasing rapidity. Cope's works stand prominent amongst all, 

 and much valuable discussion and evidence has been brought together. 



The decision, however, could only be gained by a direct study of 

 the supposed mutations, but no distinct cases of mutability were at 

 hand to provide the material. Discussions took the place of inquiry, 

 and a vast amount of literature has broadly pictured all the possi- 

 bilities and all the more or less plausible explanations without being 

 able to give proof or disproof. 



In this most discouraging state of things I concluded that the 

 only way to get out of the prevailing confusion was to return to 

 the method of direct experimental inquiry. Slow and gradual 

 changes were accepted to be invisible or nearly so; mutations, 

 however, would be clear and sharp, although of rare occurrence. 

 I determined to start on a search for them, and tried a large number 

 of species, partly native forms of my own country and partly from 

 different sources. Each of them had to be tried as to its constancy, 

 and large numbers of seedlings had to be produced and compared. 

 The chance of finding what I wanted was of course very small, and 

 consequently the number of the experiments had to be increased 

 as far as possible. 



Fortune has been propitious to me. It has brought into my 

 garden a series of mutations of the same kind as those which are 

 known to occur in horticulture, and moreover it has afforded me an 

 instance of mutability such as would be supposed to occur in nature. 

 The sudden changes, which until yet were limited to the experience 

 of the breeders, proved to be accessible to direct experimental work. 

 They can not yet in truth be produced artificially, but, on the other 

 hand, their occurrence can be predicted in some cases with enough 

 probability to justify the trial. Color changes in flowers, double 

 flowers, regular forms from labiate types, and others have been 

 produced more or less at will in my garden, and under conditions 

 which allowed of a close scientific study. The suddenness of the 

 changes and the perfection of the display of the new characters from 

 the very beginning were the most striking results. 



These facts, however, only gave an experimental proof of pheno- 

 mena which were historically known to occur in horticulture. They 

 threw light upon the way in which cultivated plants usually produce 

 new forms, but between them and the real origin of species in nature 

 the old gap evidently remained. 



24* 



