stem which could react upon the external world and modify
and combine sensations. Mill's ideas would have more apparent
connection if they could be made to correspond to 'vibratiuncles' or
physical processes of some kind. But this part of Hartley's hypothesis
had been dropped: and all reality is therefore reduced to the whirl
of vagrant and accidentally cohering ideas in brains and clusters. His
one main aim is to get rid of everything that can be called mystical
and to trace all mental processes to 'experience,' as he understands
experience--to show that we are never entitled to assert that two
ideas may not be joined in any way whatever.

The general tendency of the 'Association Philosophy' is sufficiently
clear. It may be best appreciated by comparing it to the method of the
physical sciences, which it was intended to rival. The physicist
explains the 'laws of nature' by regarding a phenomenon as due to the
varying arrangements of an indefinite multitude of uniform atoms. I
need not ask whether these atoms are to be regarded as realities, even
the sole realities, or, on the other hand, as a kind of logical
scaffolding removable when the laws are ascertained. In any case, the
assumption is necessary and most fruitful in the search for accurate
and quantitative formulae. Mill virtually assumes that the same thing
can be done by breaking up the stream of consciousness into the ideas
which correspond to the primitive atoms. What precisely these atoms
may be, how the constantly varying flow of thought can be resolved
into constituent fractions, is not easy to see. The physicist at least
supposes his atoms to have definite space relations, but there is
nothing clearly corresponding to space in the 'ideas.' They are
capable of nothing but co-existence, sequence, and likeness; but the
attempt to explain the meaning of those words ends in nothing but
repeating them. One result is the curious combination of the absolute
and the indefinitely variable. We get absolute statements because the
ultimate constituents are taken to be absolutely constant. We have
indefinite variability because they may be collocated in any
conceivable or inconceivable way. This becomes evident when we have to
do with organisms of any kind: with characters or societies an
organism varies, but varies along definite lines. But, on Mill's
showing, the organic relations correspond to the indefinitely
variable. Education is omnipotent; state constitutions ca