at a later period, made the most ample and honourable amends),
there was truth in several of his strictures on my father's treatment of
the subject; that my father's premises were really too narrow, and
included but a small number of the general truths on which, in politics,
the important consequences depend. Identity of interest between the
governing body and the community at large is not, in any practical sense
which can be attached to it, the only thing on which good government
depends; neither can this identity of interest be secured by the mere
conditions of election. I was not at all satisfied with the mode in
which my father met the criticisms of Macaulay. He did not, as I thought
he ought to have done, justify himself by saying, "I was not writing a
scientific treatise on politics, I was writing an argument for
parliamentary reform." He treated Macaulay's argument as simply
irrational; an attack upon the reasoning faculty; an example of the
saying of Hobbes, that When reason is against a man, a man will be
against reason. This made me think that there was really something more
fundamentally erroneous in my father's conception of philosophical
method, as applicable to politics, than I had hitherto supposed there
was. But I did not at first see clearly what the error might be. At last
it flashed upon me all at once in the course of other studies. In the
early part of 1830 I had begun to put on paper the ideas on Logic
(chiefly on the distinctions among Terms, and the import of
Propositions) which had been suggested and in part worked out in the
morning conversations already spoken of. Having secured these thoughts
from being lost, I pushed on into the other parts of the subject, to try
whether I could do anything further towards clearing up the theory of
logic generally. I grappled at once with the problem of Induction,
postponing that of Reasoning, on the ground that it is necessary to
obtain premises before we can reason from them. Now, Induction is mainly
a process for finding the causes of effects: and in attempting to fathom
the mode of tracing causes and effects in physical science, I soon saw
that in the more perfect of the sciences, we ascend, by generalization
from particulars, to the tendencies of causes considered singly, and
then reason downward from those separate tendencies, to the effect of
the same causes when combined. I then asked myself, what is the ultimate
analysis of this deductive process; the