188 



quality could be recognized by external characters. This led to 

 an elaborate study of all the traits and qualities of numerous 

 crops with the result that slowly a system has been established 

 for many of the grains, peas, vetches, etc., by which definite 

 qualities can be recognized from a study of their morphological 

 characters. The results of this work in the rapid improvement 

 of crops can scarcely be estimated. Success is now obtained 

 in a few years with a directness and certainty that was impossible 

 by the old method after twenty years of work. 



Naturally de Vries makes full use of this work in fortifying his 

 mutation theory, and it will be conceded that his present discus- 

 sion comes nearer to placing his theory upon an incontrovertible 

 basis than do any of his other writings. It would appear that in 

 order to establish his thesis it now remains for him to show only 

 that the continuous variations of his mutants do not overlap 

 those of the parents. 



While much that is misleading has been published about Bur- 

 bank, sufficiently accurate statements have been made to render 

 his achievements familiar to all. Even his most misguided 

 friends will recognize the scientific and appreciative consideration 

 that is given to Burbank's work. Burbank is an idealist. While 

 interested in the scientific aspects of horticulture he is primarily 

 desirous of giving to his fellow men better foods and fruits and 

 more attractive flowers, and in the cheapest form in order to dis- 

 pense their enjoyment as widely as possible. No better measure 

 of the man could be given than his dream of a spineless cactus 

 that by its adaptability to arid regions and its edible qualities 

 would make possible the doubling of the population of the world. 



Burbank's work is prosecuted on lines quite distinct from those 

 of the Swedish company. He is Uttle concerned with the im- 

 provement of a race or elementary species. His results are 

 largely gained by hybridizing. In this work he is guided by two 

 principles that are not generally tried by other breeders. He 

 does not rely upon the association of a few qualities in his hybrids, 

 but all the desirable traits possible are added at once for the pur- 

 pose of producing a chaos of forms from which valuable selec- 

 tions may be made. Thus in the California lily, Lilhivi pardali- 



