178 



hand, and the lowest probability on the other, these 

 two sorts of knowledge run into each other, and are 

 not easily to be distinguished. 



This may be illustrated by a parallel, drawn from 

 common mechanism. While you are 'offering the 

 reasons, for and against any morally-certain or pro- 

 bable proposition, imagine yourself throwing them 

 into the scales, and weighing them in a balance. If 

 the balance inclines not at all to either side, there is 

 no sort of knowledge, but downright ignorance : the 

 reasons on each side destroy each other, so that the 

 intellect cannot assent to either. And if there be 

 any decision, it is the arbitrary imposition and pre- 

 carious act of the will. If, either from its natural 

 weakness, or for want of improvement, the intellect 

 cannot find out reasons, s6 that each scale prepon- 

 derates in its turn, then it is a state of doult. If one 

 .scale preponderates but a little, and continues at a 

 ;>tay, so that the difference is barely discernible, it is 

 then only a conjecture: but if this greponderancy is 

 very plain, though there is weight enough on the 

 other side to keep the scale still prudent, then it is 

 properly probability, or opinion. When, lastly, the 

 arguments, are so strong that one of the scales weighs 

 to the ground, then it is moral certainty, and there 

 is no reasonable cause for any farther scrutiny. The 

 proposition then concludes as surely, though not so 

 necessarily, as demonstration ; which admits of no 

 weight whatever to be thrown into tho opposite 

 scale* 



