N0.55S] CAUSES AND DETERMINERS 353 



mine whether war or peace shall prevail. It might in- 

 deed be clearer if for the word determiner in such a 

 meaning, some such name as " decider" were used, but 

 it is important not to confuse a criticism of linguistic fit- 

 ness with a denial of experimental facts. All the "de- 

 terminers" spoken of in the formulations of Mendelian 

 inheritance appear clearly to be such in the sense only of 

 "deciders." 



Conclusions deducible only from discovery of all the 

 conditions necessary to produce a certain result must, of 

 course, not be drawn from experiments showing a de- 

 terminer only in the sense of "decider" between two 

 possibilities; this appears not infrequent. Such illegiti- 

 mate conclusions are perhaps most usually drawn when 

 persons understanding determination in the first sense 

 examine the statements of those that use the word in the 

 second sense; this is a source of polemics. 



Since to produce almost any result an indefinitely great 

 number of preceding conditions, of diverse sorts, must 

 have been fulfilled, and since neither thought nor prac- 

 tical investigation can handle all these at once, it be- 

 comes necessary to so analyze our problems that at a 

 particular juncture only one cause or determiner (and 

 that a definite one) need be sought. The key for this is 

 the following principle : 



A single sufficient determining factor can be found 

 only for the difference between two cases. 



With relation to this, several points must be grasped. 



1. Evidently two cases may be so chosen that the dif- 

 ference between them is not due to a single determining 

 cause. But by proper analysis problems can be brought 

 (at least usually) to a situation where but a single de- 

 termining cause is required; this is done by comparing 

 cases that differ only in certain defined features ; and in 

 bringing the two cases closer and closer together, till 

 finally the difference between them is due to but a single 

 experimental cause. 



2. For the difference between two cases that are di- 



