example of an art more delicate than Scott elsewhere exhibits.

One does not know what Professor Child would have said to my arguments.
He never gave a criticism in detail of the ballad and of the
circumstances in which Scott acquired it. A man most reasonable, most
open to conviction, he would, I think, have confessed his perplexity.

Scott did not interpolate a single stanza, even where, as Hogg wrote,
he suspected a lacuna in the text. He neither cut out nor improved the
cryingly modern stanzas. He kept them, as he kept several stanzas in
Tamlane, which, so he told Laidlaw, were obviously recent, but were in
a copy which he procured through Lady Dalkeith. {51a}

By neither adding to nor subtracting from his MS. copy of Auld
Maitland, Scott proved, I think, his respect for a poem which, in its
primal form, he believed to be very ancient. We know, at all events,
that ballads on the Maitland heroes were current about 1580. So, late
in the sixteenth century, were the ballads quoted by Hume of Godscroft,
on the murder of the Knight of Liddesdale (1354), the murder of the
young Earl of Douglas in Edinburgh Castle (1440), and the battle of
Otterburn. Of these three, only Otterburne was recovered by Herd,
published in 1776. The other two are lost; and there is no prima facie
reason why a Maitland ballad, of the sort current in 1580, should not,
in favourable circumstances, have survived till 1802.

As regards the Shepherd's ideas of honesty in ballad-collecting at this
early period, I have quoted his letter to Laidlaw of 20th July 1802.

Again, in the case of his text from recitation of the Ballad of
Otterburne (published by Scott in The Minstrelsy of 1806), he gave the