154 THE ROYAL SOCIETY OF CANADA 
accurately and deeply into the record; and the individuality therein 
imprinted may be studied and analysed with a care and exactness 
impossible in the fleeting utterances of unrecorded conversation. For 
this reason it is possible to determine from the record alone the accuracy 
and fidelity of the writer in exemplifying the five requisites and in apply- 
ing the corresponding correct processes for formulating trustworthy 
statements. From the record alone it is even possible to tell whether 
he applies these processes with varying accuracy and faithfulness. 
Thus the writer of a record may be impartial except where he deals with a 
certain individual toward whom he has developed an antipathy, or when 
he touches on a certain question in which he is prejudiced. He cannot, 
however, let such an antipathy or prejudice seriously affect his conclu- 
sions without a corresponding effect appearing in his narrative; and in 
this case the record, accepted as trustworthy in other respects, will be 
rejected as untrustworthy wherever this antipathy or prejudice appears. 
So also with respect to the other four processes, if a writer is generally 
intelligent and clear in the preparation and statement of his record, 
but in some points the reverse; if he, for the most part, gives only 
concrete information and yet, by way of exception, indulges occasionally 
in philosophizing or a jest; if from a proper poise ordinarily well main- 
tained he lapses casually into exaggeration; or if from time to time 
he inserts in the narrative admittedly unnecessary conclusions such as 
usually he excludes: there cannot be at any point a serious departure 
of this sort from his ordinary standard of trustworthiness without 
evidence of it in the narrative; and at that point the record, elsewhere 
accepted as trustworthy, will be rejected as untrustworthy. By the 
narrative itself, therefore, not only as a whole but in respect of its 
various parts, if between them there be a difference, the trustworthiness 
of the record is to be judged; and in the decision the identity and 
reputation of the writer can weigh little. For a veteran historian, well 
known and having earned a deserved reputation for trustworthiness 
by works written in his prime, may in his old age produce a narrative 
of inferior value. Such a narrative will bear in itself the mark of his 
decaying powers, and its trustworthiness will be estimated according 
to its own worth, not according to his reputation. Thus the question 
of the respective trustworthiness not only of different writers, but of 
different works by the same writer and even of different parts in the 
same work, is of necessity determined by the exemplification of the 
requisites for trustworthiness in the narrative itself; and in determining 
the trustworthiness of a record as well as of any of its parts, this exempli- 
fication constitutes not inferior evidence, to be resorted to only in heu 
of nothing better: it is the first, as well as the ultimate, and only decisive, 
test. 
