is
reply that his meaning was the same throughout, that he knew all the
while Irenaeus must be quoting from some one else, and that he 'did what
was possible to attract attention to the actual indirect construction.'
[127:2] Why then did he translate the oblique construction as if it were
direct? Why, after quoting as parallels a number of direct sentences in
Irenaeus containing quotations, did he add, 'These are all direct
quotations by Irenaeus, as is most certainly that which we are now
considering, which is introduced in precisely the same way?' [127:3] Why
in his fourth edition, in which he first introduces a recognition of the
oblique construction, did he withdraw all these supposed parallels,
which, if his opinion was unchanged, still remained as good for his
purpose (whatever that purpose might be) as they had ever been? Further
discussion on this point would obviously be wasted. I can only ask any
reader who is interested in this matter to refer to the book itself, and
more especially to compare the fourth [128:1] with the earlier editions,
that he may judge for himself whether any other interpretation, except
that which I and others besides myself [128:2] have put upon his words,
was natural. The author has declared his meaning, but I could only judge
by his language.

I now proceed to notice some other of the chief points in our author's
reply; and perhaps it may be convenient in doing so to follow the order
adopted in my original article to which it is a rejoinder.

1. In the first place then, the author is annoyed that I spoke
disparagingly of his scholarship [128:3]; and in reply he says that the
criticism in which I have indulged 'scarcely rises above the correction
of an exercise or the conjugation of a verb.' [128:4] I cannot help
thinking this language unfortunate from his own point of view; but let
that pass. If the reader will have the goodness to refer back to my
article, he will find that, so far from occupying the main part of it on
points of scholarship which have no bearing on the questions under
discussion, as the author seems to hint, I have taken up about
two-thirds of a page only [128:5] with such matters. In the other
instances which I have selected, his errors directly affect the argument
for the time being at some vital point. It would have been possible to
multiply examples, if examples had been needed. I might have quoted, for
instance, such renderings as [Greek: katabas peripateito] '