rce and range. But there is still an interval within which only
approximate conclusions are possible. These conclusions need to be
supplemented from the phenomena of the documents themselves. In
the relation of the Gospels to the growth of the Christian society
and the development of Christian doctrine, and especially to the
great turning-point in the history, the taking of Jerusalem, there
is very considerable internal evidence for determining the date
within which they must have been composed. It is well known that
many critics, without any apologetic object, have found a more or
less exact criterion in the eschatological discourses (Matt. xxiv,
Mark xiii, Luke xxi. 5-36), and to this large additions may be
made. As I hope some day to have an opportunity of discussing the
whole question of the origin and composition of the Synoptic
Gospels, I shall not go into this at present: but in the mean time
it should be remembered that all these further questions lie in
the background, and that in tracing the formation of the Canon of
the Gospels the whole of the evidence for miracles--even from this
_ab extra_ point of view--is very far from being exhausted.

There is yet another remaining reason which makes the present
enquiry of less importance than might be supposed, derived from
the particular way in which the author has dealt with this
external evidence. In order to explain the _prima facie_
evidence for our canonical Gospels, he has been compelled to
assume the existence of other documents containing, so far as
appears, the same or very similar matter. In other words, instead
of four Gospels he would give us five or six or seven. I do not
know that, merely as a matter of policy, and for apologetic
purposes only, the best way to refute his conclusion would not be
to admit his premisses and to insist upon the multiplication of
the evidence for the facts of the Gospel history which his
argument would seem to involve. I mention this however, not with
any such object, but rather to show that the truth of Christianity
is not intimately affected, and that there are no such great
reasons for partiality on one side or on the other.

I confess that it was a relief to me when I found that this must
be the case. I do not think the time has come when the central
question can be approached with any safety. Rough and ready
methods (such as I am afraid I must call the first part of
'Supernatural Religion') may indeed cut the Gordian