660 PSYCHOLOGY. 



which we have considered, in which the recurring relation 

 is either of difference, of likeness, of kind, of numerical ad- 

 dition, or of prolongation in the same linear or plane direc- 

 tion. It is therefore not a purely formal law of thinking, 

 but flows from the nature of the matters thought about. It 

 will not do to say universally that in all series of homo- 

 geneously related terms the remote members are related to 

 each other as the near ones are ; for that will often be 

 untrue. The series A is not B is not C is not D . o . does 

 not permit the relation to be traced between remote terms. 

 From two negations no inference can be drawn. Nor, to 

 become more concrete, does the lover of a woman generally 

 love her beloved, or the contradictor of a contradictor con- 

 tradict whomever he contradicts. The slayer of a slayer 

 does not slay the latter's victim ; the acquaintances or ene- 

 mies of a man need not be each other's acquaintances or 

 enemies ; nor are two things which are on toj3 of a third 

 thing necessarily on top of each other. 



All skipping of intermediaries and transfer of relations 

 occurs within homogeneous series. But not all homoge- 

 neous series allow of intermediaries being skipped and re- 

 lations transferred. It depends on which series they are, 

 on what relations they contain,* Let it not be said that it 

 is a mere matter of verbal association, due to the fact that 

 language sometimes permits us to transfer the name of a 

 relation over skipped intermediaries, and sometimes does 

 not ; as where we call men ' progenitors ' of their remote as 

 well as of their immediate posterity, but refuse to call them 

 ' fathers ' thereof. There are relations which are intrinsi- 

 cally transferable, whilst others are not. The relation of 

 condition, e.g., is intrinsically transferable. What conditions 

 a condition conditions what it conditions — " cause of cause 

 is cause of effect." The relations of negation and fr^istratioUf 

 on the other hand, are not transferable : what frustrates a 

 frustration does not frustrate what it frustrates. No 

 changes of terminology would annul the intimate difference 

 between these two cases. 



* Cf . A. de Morgan : Syllabus of a proposed System of Logic (1860), pp. 

 46-56. 



