[ 8S ] 



We can alfo from thefe principles fhevv why after a certain 

 number of witneffes have attelled a fad, any farther evidence is 

 fuperfluous. 



These principles likew^ife, as Dr. Waring obferves, may be 

 applied to the inveftigation of the probability of the truth of the 

 decifion by any number of voters, and many other cafes ; the 

 probability of each voter voting truly being fuppofed given. But, 

 as he alfo obferves, it is impoilible to determine ihe knowledge, 

 integrity, and various influences which adtuate each perfon, and 

 confequently to determine the probability of their vocing truly. 



But though we cannot determine the adual probability, yet 

 fmce the voters are to be fuppofed of equal integrity, kn^vvledge, 

 &c. we can determine the relative probabilities of ^he iruth of 

 the decifions by different majorities ; and on thefe principles 

 Mens. Condorcet has enquired into the laws according to which 

 the majorities, which decide queftions in deliberative affemblies, 

 aught to be regulated. 



Thus fuppofe the enading of a new law were propofed to a 

 deliberative affembly, fuch a majority fhould be required as would 

 give a very great probability of the juftice of their deci:l<^n; for 

 it is much better that no law fhould be enaded than a bad one. 

 A majority of more than one fingle voice feems alfo requifite ia 

 fome qucflions of a civil nature, as for inftance in long continued 



poffeilion 



