orms, it
is indeterminate. It acts in one sort of way under the persuasion of the
sheep-form, and in another sort of way under the persuasion of the [16]
dog-form, and we cannot tell how it will act until we know which form is
going to capture it. No amount of study bestowed on the common material
nature will enable us to judge how it will behave under the persuasion of
the higher organizing form. The only way to discover that is to examine the
higher form itself.

Every form, then, will really be the object of a distinct science. The form
of the sheep and the form of the dog have much in common, but that merely
happens to be so; we cannot depend upon it, or risk inferences from sheep
to dog: we must examine each in itself; we shall really need a science of
probatology about sheep, and cynology about dogs. Again, the common
material nature has its own principles of being and action, so it will need
a science of itself, which we may call hylology. Each of these sciences is
mistress in her own province; but how many there are, and how puzzlingly
they overlap! So long as we remain within the province of a single science,
we may be able to think rigorously, everything will be 'tight'. But as soon
as we consider border-issues between one province and another, farewell to
exactitude: everything will be 'loose'. We can think out hylology till we
are blue in the face, but we shall never discover anything about the entry
of material elements into higher organizations, or how they behave when
they get there. We may form perfect definitions and descriptions of the
form of the dog as such, and still derive no rules for telling what
elements of matter will enter into the body of a given dog or how they will
be placed when they do. All we can be sure of is, that the dog-form will
keep itself going in, and by means of, the material it embodies--unless the
dog dies. But what happens to the matter in the body of the dog is
'accidental' to the nature of the matter; and the use of this matter,
rather than of some other equally suitable, is accidental to the nature of
the dog.

No account of material events can dispense with accidental relations
altogether. We must at least recognize that there are accidental relations
between particular things. Accident in the sense of brute fact had to be
acknowledged even by the tidiest and most dogmatic atomism of the last
century. That atomism must allow it to be accidental, in this sense, that
t