OTHER THEORIES OF SPECIES-FORMING. 311 



a new and pleasing odour it may often be sufficient simply to lose 

 one bad element in an old odour. So one might go on for some 

 pages with specific conclusions or deductions reached by Burbank 

 on a basis of experience. But it is true that he has at his command 

 the knowledge of no new fundamental scientific principles to give him 

 advantage over us. And yet none of us has done what Burbank 

 has been able to do. although many of us have tried. What then 

 is it that Burbank brings to his work of modifying organisms 

 swiftly and extremely and definitely that others do not? 



"To answer this it will be advisable to analyse in general terms, 

 at least, the various processes which either singly, or in combina- 

 tions of two or three, or all together, are used by Mr. Burbank in 

 his work. We may roughly classify these processes and means. 

 First, there is the importation from foreign countries, through 

 many correspondents, of a host of various kinds of plants, some 

 of economic value in their native land and some not, any of which 

 grown under different conditions here may prove specially vigorous 

 or prolific or hardy, or show other desirable changes or new quali- 

 ties. Among these importations are often special kinds particularly 

 sought for by Burbank to use in his multiple hybridisations ; kinds 

 closely related to our native or to already cultivated races which, 

 despite many worthless characteristics, may possess one or more 

 particular, valuable ones needed to be added to a race already useful 

 to make it more useful. Such an addition makes a new race. 



"Second, the production of variations, abundant and extreme, by 

 various methods, as (a) the growing under new and. usually, more 

 favourable environment (food supply, water, temperature, light, 

 space, etc.) of various wild or cultivated forms, and (b) by hybridi- 

 sations between forms closely related, less closely related and, 

 finally, as dissimilar as may be (not producing sterility), this 

 hybridising being often immensely complicated by multiplying 

 crosses, i. e., the offspring from one cross being immediately crossed 

 with a third form, and the offspring of this with still another form, 

 and so on. These hybridisations are made sometimes with very 

 little reference to the actual useful or non-useful characteristics of 

 the crossed parents, with the primary intention of producing an 

 unsettling or instability in the heredity, of causing, as Burbank 

 sometimes says, 'perturbations' in the plants, so as to get just as 

 wide and as large variation as possible. Other crosses are made, 

 of course, in the deliberate attempt to blend, to mix, to add together, 

 two desirable characteristics, each possessed by only one of the 

 crossed forms. Some crosses are made in the attempt to extin- 

 guish an undesirable characteristic. 



"Third, there is always immediately following the unusual produc- 



