[ 45 1 



history that, in old times, men hardly ever went out 

 of their way to indicate the line of thought which J 

 led them to make any particular historical state- 

 ment; we can sometimes supply the omission by 

 conjecture, but the value of the historical evidence in 

 this case is much less than it would have been if the 

 syntax had been plainly indicated by the original 

 writer. Again, men seem especially apt to forget any 

 syntaxes which they have not been ab'e to bring into 

 harmon} T with their main conclusions, and it is this 

 forgetfulness which constitutes most of the bias often 

 recognisable in evidence otherwise trustworthy. 



The first question to be investigated regarding 

 the syntaxes involved in evidence is whether they 

 are the best which that particular witness could make, 

 which is equivalent to an inquiry whether the 

 evidence was given in good faith. It is recognised 

 that the things to be considered here are any 

 circumstances which might have induced a desire to 

 mislead, any indications of the character of the 

 witness, and any examples of the kind of evidence 

 given by him in other matters, especially in similar 

 matters. Besides these, account should be taken of 

 the kind of syntaxis which is usual among men in an 

 environment similar to that of the witness. Here, 

 again, the scientific criterion is applied by forming the 

 clearest and most direct syntax which can be made to 

 include all the available reminiscences bearing upon 

 these points. 



