91

1908 Nov. 19
Logic
I. i. 10

of my italicized sentence had merely been an ingenious hypothe-
sis framed to account for the strange historical series of coincidences,
the latter (as I shall show you later) would have furnished no support
for belief in the former: but in point of fact it was quite the other way.
The former was an induction by which my life-long intimacy with
many scientific men had led me to believe that what I had remarked
of my own quite hap-hazard competence to attack problems was true
of men of science generally, – a belief that had some additional
support. Still, though that induction was quite legitimate,
I remarked two elements of weakness in it, first, that it was of that
crude kind to which one ought not to trust exclusively when one can
avoid doing so, and secondly that the observations had not been regular-
ly recorded but had been preserved only by a treacherous memory that [??]