ally a preface to the six selected lives,
which he edited for Messrs Macmillan in 1878) is a most excellent
piece of work. His selection of the Lives is perhaps not quite
unerring. For he ought surely to have given the "Cowley," with its
(from his own point of view) invaluable _point de repere_ in the
estimate of the "metaphysicals." And he might have missed the "Swift,"
which, though extremely interesting as a personal study from its
mixture of prejudice and constraint, its willingness to wound, and
yet--not its fear but--its honest compunction at striking, is, for the
purpose of the volume, misplaced. But he had a right to give what he
chose: and his preface has points of the very highest value. The
opening passage about the _point de repere_ itself, the fixed
halting-place to which we can always resort for fresh starts, fresh
calculations, is one of the great critical _loci_ of the world,
and especially involves the main contribution of the nineteenth
century to criticism if not to literature altogether. We may exalt,
without very much doubt or dread, the positive achievements of the
century of Tennyson and Browning, of Carlyle and Thackeray, of Heine
and Hugo. But we have seen such strange revolutions in this respect
that it may not do to be too confident. The glory of which no man can
deprive our poor dying _siecle_ is that not one, of all the
others since history began, has taken such pains to understand those
before it, has, in other words, so discovered and so utilised the
value of _points de repere_. It may be that this value is, except
in the rarest cases, all that a critic can ever pretend to--that he
may be happy if, as few do, he reaches this. But in the formulation of
the idea (for he did much more than merely borrow it from the French)
Mr Arnold showed his genius, his faculty of putting

"What oft was thought but ne'er so well expressed."

And when a man does this in prose or in verse, in criticism or in
creation, he has his reward--a reward that no man can take away, even
if any one were disposed to try.

As a whole, _Mixed Essays_ itself, which followed _Last Essays
on Church and Religion_ at an interval of two years, is an almost
immeasurably livelier book than its predecessor, and to some judgments
at least seems to excel that predecessor in solid value as much as in
the graces. "Mixed" is perhaps not a strictly accurate title, for the
volume consists of two halves, the contents of each of which