of the brain may be as certain as the association of light with the
rising of the sun. But whereas in the latter case we have unbroken
mechanical connection between the sun and our organs, in the former
case logical continuity disappears. Between molecular mechanics and
consciousness is interposed a fissure over which the ladder of
physical reasoning is incompetent to carry us. We must, therefore,
accept the observed association as an empirical fact, without being
able to bring it under the yoke of _a priori_ deduction.

*****

Such were the ponderings which ran habitually through my mind in the
days of my scientific youth. They illustrate two things--a
determination to push physical considerations to their utmost
legitimate limit; and an acknowledgment that physical considerations
do not lead to the final explanation of all that we feel and know.
This acknowledgment, be it said in passing, was by no means made with
the view of providing room for the play of considerations other than
physical. The same intellectual duality, if I may use the phrase,
manifests itself in the following extract from an article entitled
'Physics and Metaphysics,' published in the 'Saturday Review' for
August 4, 1860:

'The philosophy of the future will assuredly take more account than
that of the past of the dependence of thought and feeling on physical
processes; and it may be that the qualities of the mind will be
studied through organic combinations as we now study the character of
a force through the affections of ordinary matter. We believe that
every thought and every feeling has its definite mechanical
correlative--that it is accompanied by a certain breaking up and
remarshalling of the atoms of the brain. This latter process is
purely physical; and were the faculties we now possess sufficiently
expanded, without the creation of any new faculty, it would doubtless
be within the range of our augmented powers to infer from the
molecular state of the brain the character of the thought acting on
it, and, conversely, to infer from the thought the exact molecular
condition of the brain. We do not say--and this, as will be seen, is
all-important--that the inference here referred to would be an _a
priori_ one. But by observing, with the faculties we assume, the state
of the brain and the associated mental affections, both might be so
tabulated side by side that, if one were given, a mere reference to
the table would declare the ot