[ 201 J 



flrative certainty is unattainable, and in not pointing out tlic 

 caufe which in thofe propofitions renders certainty hopelefs. 

 He alfo appears to have been erroneous in his favourite poli- 

 tion, that moral truths are capable of flricl demonflration. 



To fupply this want of diftindlnefs, and to corredl thefe 

 errors, I would divide all our enquiries into three ckiTes ; the 

 firft of which Ihould comprehend all thofe in which our ideas 

 are compared together, without being confidered as connedled 

 by the relation of caufe and efFefl, but merely regarded as 

 independent objects of thought, correfponding in fome parti- 

 culars which enter into the compofition of each idea ; the 

 fecond fhould include thofe in which a confideration. of the 

 relation of caufe and efFedt is diredlly or indirecflly involved, 

 limited however to the mere fadl of their connedtion, and not 

 extending to the nature of that connedlion or manner of operation ; 

 and the third Ihould confift of enquiries into the nature of 

 caufes or che manner of operation. 



By the word caufes I mean not only phyfical but moral 

 caufes, without pretending to determine the quantity of the 

 influence of the latter. If motives are allowed to have any 

 influence, it is fufiicient to entitle them to be placed amongft 

 caufes. Of thofe three clafl^es the firfl; appears to me to be 

 the region of demonftrative certainty, the fecond to be that 

 in which probability alone is attainable, and the third to be 

 that of abfolute ignorance. In this divilion they have been. 



Vol. V. Co arranged. 



