BURBANK'S HORTICULTURAL NOVELTIES 221 



caution is to be observed in the scientific interpretation of the 

 breeder's results. 



As a general rule, we may state that the broad hues of 

 practical hybridizing and selection afford highly valuable 

 resources for theoretical discussions, but that on single 

 points they should not be accepted as definite proofs, but 

 only as indications for more sharply circumscribed experi- 

 ments. Even with this restriction, however, the value of 

 Burbank's work for the doctrine of evolution, compels our 

 highest admiration. 



D. I^IUTATIONS IN HORTICULTURE 

 For the larger number of horticultural varieties we do 

 not know the origin. jNIany of them are older than the his- 

 torical records, others have been found in the wild state or 

 in foreign countries or in old-fashioned gardens. A certain 

 number appeared in nurseries and cultures, but their first 

 generations were overlooked and they were appreciated only 

 when it was already too late to study their first appearance. 

 Whenever this first appearance has been noted and re- 

 corded the development of the variety took one or two or a 

 small number of years. Sometimes it appeared with the 

 full chsplay of its quahties and proved constant from seed 

 from the very beginning. But such cases are rare. Ordi- 

 narily it was discovered, at first, in a very imperfect and often 

 scarcely perceptible degree of development, or it was not 

 constant from seed. In both cases it had to be developed 

 by selection in order to insure a normal degree of purity and 

 constancy. This process requires some years, and their 

 number is different for different species, according to their 

 capacity for self-fertilization, their fertility, and other factors. 

 Whenever the varietal mark is plain in the first year, 

 selection has only to produce constancy. Tliis might be 

 attained at once, if the visits of insects could be excluded, 



