[ 94 ] 



logarithms encreafe nearly in the fame proportion with the ab- 

 folute numbers, ic follows, that the probability that a future 

 event will happen in the fame manner as in previous trials, will 

 be nearly proportional to the number of thefe previous ex- 

 periments. 



By this it is demonftrable, that the experience of the paft 

 is a principle of probability for the future ; and that the more 

 frequently we have experienced an event to happen, the greater 

 reafon have we to expedl, that it will happen in the next trial. 

 Now fmce in order that we fliould have a given degree of pro- 

 bability for the events happening in a particular way, a certain 

 number of experiments is requifite, it follows converfely, that a 

 given number of experiments will produce fome determinate 

 degree of probability. This probability, we may perceive, de- 

 pends merely on the number of experiments ; fo that, this 

 number being the fame, the degree of probability will be the 

 fame whatever be the fpecific nature of the event. 



Hence therefore, fince the inference which we make with 

 refpedl to the mechanical phaenomena of nature, as well as with 

 refpedl to the veracity of human teftimony, are both equally 

 derived from experience of the pafl: ; if the teftimony be of fuch 

 a nature as has never deceived us, the probabilities in both cafes 

 will bear to each other fome determinate ratio, which, however 

 Ignorant we may be of the pr' iciples of the calculation, depends 



on 



