nty-five years afterwards; but they ought manifestly to retain their
place, as arranged by Wordsworth in the edition of 1845. The case is
much the same with regard to the "Duddon Sonnets." They were first
published in 1820: but No. xiv. beginning:

O mountain Stream! the Shepherd and his Cot,

was written in the year 1806, and appears in the edition of 1807. This
sonnet will be printed in the series to which it belongs, and not in its
chronological place. I think it would be equally unjust to remove it
from the group--in which it helps to form a unity--and to print it twice
over. [3] On the other hand, the series of "Poems composed during a Tour
in Scotland, and on the English Border, in the Autumn of 1831"--and
first published in the year 1835, in the volume entitled "Yarrow
Revisited, and Other Poems"--contains two, which Wordsworth himself
tells us were composed earlier; and there is no reason why these poems
should not be restored to their chronological place. The series of
itinerary sonnets, published along with them in the Yarrow volume of
1835, is the record of another Scottish tour, taken in the year 1833;
and Wordsworth says of them that they were "composed 'or suggested'
during a tour in the summer of 1833." We cannot now discover which of
them were written during the tour, and which at Rydal Mount after his
return; but it is obvious that they should be printed in the order in
which they were left by him, in 1835. It may be noted that almost all
the "Evening Voluntaries" belong to these years--1832 to 1835--when the
author was from sixty-two to sixty-five years of age.

Wordsworth's habit of revision may perhaps explain the mistakes into
which he occasionally fell as to the dates of his Poems, and the
difficulty of reconciling what he says, as to the year of composition,
with the date assigned by his sister in her Journal. When he says
"written in 1801, or 1802," he may be referring to the last revision
which he gave to his work. Certain it is, however, that he sometimes
gave a date for the composition, which was subsequent to the publication
of the poem in question.

In the case of those poems to which no date was attached, I have tried
to find a clue by which to fix an approximate one. Obviously, it would
not do to place all the undated poems in a class by themselves. Such an
arrangement would be thoroughly artificial; and, while we are in many
instances left to conjecture, we can always say that such a