3 io PLANT-BREEDING 



plants should be laid in their hands only. We wish to know 

 how they work, and how their great achievements are 

 obtained. We wish to study the rules and laws underlying 

 their attainments, in order to apply them to as many instances 

 as possible. 



In science, on the other hand, the innumerable cases of 

 observed correlations lead to the question of the more inti- 

 mate causes of this phenomenon. Our imagination cannot 

 be content with the outward features of the facts; we wish 

 to have some idea of their innermost nature. In this line of 

 thought the principal difficulty lies in the absence of definite- 

 ness in the objects we have to deal with. We speak of 

 correlation and interdependency, but we have no idea as 

 to what the things are that should be related to or depend- 

 ent upon each other. 



Here the idea suggests itself that in order to be corre- 

 lated the characters must begin by being independent entities, 

 which through some later means may come into relation with 

 others. Perhaps this may not be the real way in which 

 nature proceeds, but at all events it is the way in which we 

 should proceed in our analysis. Thus we come to the con- 

 ception of units which govern and control the visible char- 

 acters and qualities. 



A new scope for investigation is opened by this concep- 

 tion. Each organism appears to us as a microcosm, con- 

 sisting of thousands of elementary entities, which combine 

 to give it its form and functions. The study of these units 

 encroaches upon systematic and comparative sciences, as 

 well as upon the investigations into the physiology and the 

 evolution of all living beings. All comparison culminates 

 in the question as to which units are common to. the species 

 under discussion, and which units are the causes of their 

 differences. Systematic affinity is reduced to the same 

 principle. It is founded on the community of a more or 



