lfilments too would occupy their modest
position in the rank and file of marching existence. To omniscience the
idea of cause and effect would be unthinkable. If all things were
perceived together and co-existed for thought, as they actually flow
through being, on one flat phenomenal level, what sense would there be
in saying that one element had compelled another to appear? The relation
of cause is an instrument necessary to thought only when thought is
guided by presumption. We say, "If this thing had happened, that other
thing would have followed"--a hypothesis which would lapse and become
unmeaning had we always known all the facts. For no supposition contrary
to fact would then have entered discourse.

[Sidenote: Indirect methods of attaining it.]

This ideal of direct omniscience is, however, impossible to attain; not
merely accidental frailties, but the very nature of things stands in the
way. Experience cannot be suspended or sustained in being, because its
very nucleus is mobile and in shifting cannot retain its past phases
bodily, but only at best some trace or representation of them. Memory
itself is an expedient by which what is hopelessly lost in its totality
may at least be partly kept in its beauty or significance; and
experience can be enlarged in no other way than by carrying into the
moving present the lesson and transmitted habit of much that is past.
History is naturally reduced to similar indirect methods of recovering
what has lapsed. The historian's object may be to bring the past again
before the mind in all its living reality, but in pursuing that object
he is obliged to appeal to inference, to generalisation, and to dramatic
fancy. We may conveniently distinguish in history, as it is perforce
written by men, three distinct elements, which we may call historical
investigation, historical theory, and historical romance.

[Sidenote: Historical research a part of physics.]

Historical investigation is the natural science of the past. The
circumstance that its documents are usually literary may somewhat
disguise the physical character and the physical principles of this
science; but when a man wishes to discover what really happened at a
given moment, even if the event were somebody's thought; he has to read
his sources, not for what they say, but for what they imply. In other
words, the witnesses cannot be allowed merely to speak for themselves,
after the gossiping fashion familiar in Herodotus;