552 j. M. KIGG :

the theory would thus be saved at the expense of its signifi- cance. 1 If, however, the idea of unconditionality be abandoned and causation be defined as conditionally invariable antecedence and sequence between phenomena, then the difficulty pre- sented by the so-called ' plurality of causes ' has to be met. If an event may supervene upon any one of several independent antecedents, in what sense can any one of such antecedents be said to be the invariable antecedent of that effect ? In the sense, it may perhaps be urged, that, given any one of them, the effect invariably follows. This, however, is to abandon uniformity of antecedence, reducing causation to mere uniformity of sequence ; it is necessary to constitute even a conditionally invariable antecedent, not merely that the sequent should invariably follow upon it, but that in de- fault of the antecedent happening the sequent should not happen. If there are several possible antecedents of a given event, no one of them can answer the description of even a conditionally invariable antecedent. Heat, e.g., may be generated in a variety of ways. The Method of agreement enables us to say that, given one or other of certain antece- dents, heat will always, in the absence of counteracting causes, be generated. The method of difference establishes that, in the absence of one or other of the said antecedents, heat will not be generated ; but it does not entitle us to say that any one of these antecedents is even the conditionally invariable antecedent of the generation of heat ; for, though the conditions remain otherwise strictly identical, heat may be generated either by friction, or percussion, or electricity. The method of difference is powerless to eliminate the plurality of causes ; that can only be done by hypothesis by assuming, e.g., in the instance of heat, that at bottom the various modes in which it is generated are identical as being all modes of motion. Moreover, even in cases into which the plurality of causes does not enter, it cannot without absurdity be maintained that causation is mere conditional invariability of antece- dence and sequence. On such a theory high tide would have to be reckoned the cause of low tide, and vice versa. If then, discarding the idea that either unconditionality or antecedence 1 " That which will be followed by a given consequent, when and only when some third circumstance also exists, is not the cause, even though no case should ever have occurred in which the phenomenon took place with- out it " (Mill's Logic, 8th ed., bk. iii., c. 5, 6). It is safe to say that no observable event is ever followed by another except conditionally upon some third circumstance being given.