[ ^02 ] 



But the conviQion produceJ by tefliniony is capable of being 

 ■carried much higher than the convi6lion' produced by other 

 experience ; and the reafon is this, becaufe there may be con- 

 current telVmonies with refpecSl to the truth of the fame indi- 

 vidual fa<5l, whereas there can be no concurrent experiments 

 with refpedl to an individual experiment. There may indeed 

 be analogous experiments, in the fame manner as there may be 

 analogous teftimonies ; but in a courfe of nature there is but 

 one continued feries of events, whereas in teftimony, fince the 

 fame event may be obferved by different witnefTes, their concur- 

 rence is capable of producing a convidlion more cogent than 

 any which is derived from any other fpecies of events in the 

 courfe of nature. In material pha^nomena, the probability of an 

 expelled event depends folely on analogous experiments, which 

 have been made previous to the event ; and this probability admits 

 of indefinite encreafe from the unlimited encreafe of the number of 

 thefe precedent experiments. The credibility of a witnefs arifes 

 likewife from our experience of the veracity of previous witnefTes, 

 and admits of unlimited encreafe, according to their number; 

 and the law of its encreafe is, of courfe, the fame with that de- 

 rived from phyfical events. There is however another fource of 

 the encreale of tef\imony, which is likewife unlimited, derived 

 from the number of concurrent witnefTes, and its encreafe on this 

 account follows a law different from the former. The evidence 

 .of teftimony therefore admitting of an unlimited encreafe on two 

 different accounts, and the probability of the happening of any 



fpecific 



