BH. xvn.] SOCIOLOGY AND FREE-WILL. 181 



the worst enemy tlian the best friend of the murdered man. 

 If we see a man jump fi'oni a fourth-story window, we must 

 licware of too hastily inferring his insanity, since he may be 

 merely exercising his free-will ; the intense love of life im- 

 planted in the human breast being, as it seems, unconnected 

 with attempts at suicide or at self-preservation. We can thus 

 frame no theory of human actions whatever. The countless 

 empirical maxims of every-day life, the embodiment as they 

 are of the inherited and organized sagacity of many genera- 

 tions, become wholly incompetent to guide us ; and nothing 

 which any one may do, ought ever to occasion surprise. The 

 mother may strangle her first-born child, the miser may cast 

 his long-treasured gold into the sea, the sculptor may break 

 in pieces his lately-finished statue, in -the presence of no 

 other feelings than those which before led them to cherish, 

 to hoard, and to create. 



To state these conclusions is to refute their premise. 

 Probably no defender of the doctrine of free-will could be 

 induced to accept them, even to save the theorem with which 

 they are inseparably wrapped-up. Yet the dilemma cannot 

 be avoided. Volitions are either caused, or they are not. If 

 they are not caused, an inexorable logic brings us to the 

 absurdities just mentioned. If they are caused, the free-will 

 doctrine is annihilated. No help is afforded by the gratuitous 

 hypothesis tliat there is a connection between the act and the 

 motive, which yet is not a causal connection. Such con- 

 nection, if it exist, must be a case either of conditional 

 invariable sequence, or of unconditional invariable sequence. 

 On the first supposition, we have a case like the succession of 

 day and night, in which both terms of the sequence are 

 conditioned upon a third fact ; so that here we do not escape 

 causation. The second supposition is but an asssrtion of 

 causation in otlier words. While to take refuge in the 

 postulate that this assumed connection is a case of variable 

 sequence, is to af&rm and deny connection in the same breath, 



