says: 'The appeal to the sailors shows that Mary Hamilton dies in a
foreign land--not that of her ancestors.' Yet the ballad makes her
die in or near the Canongate! Moreover, the family of the Mary
Hamilton of 1719 had been settled in Russia for generations, and
were reckoned of the Russian noblesse. The verses, therefore, on
either theory, are probably out of place, and are perhaps an
interpolation suggested to some reciter (they only occur in some of
the many versions) by a passage in 'The Twa Brithers.'*

*Child, i. 439.

We now reach the most important argument for the antiquity of 'The
Queen's Marie.' Mr. Courthope has theoretically introduced as
existing in, or after, 1719, 'makers' who could imitate to deception
the old ballad style. Now Maidment remarks that 'this ballad was
popular in Galloway, Selkirkshire, Lanarkshire, and Aberdeen, AND
THE VERY STRIKING DISCREPANCIES GO FAR TO REMOVE EVERY SUSPICION OF
FABRICATION.' Chambers uses (1829) against Sharpe the same argument
of 'universal diffusion in Scotland.' Neither Mr. Child nor Mr.
Courthope draws the obvious inferences from the extraordinary
discrepancies in the eighteen variants. Such essential
discrepancies surely speak of a long period of oral recitation by
men or women accustomed to interpolate, alter, and add, in the true
old ballad manner. Did such rhapsodists exist after 1719? Old
Charlie, for one, did not sing or sell the old ballads. Again, if
the ballad (as it probably would be in 1719) was PRINTED, or even if
it was not, could the variations have been evolved between 1719 and
1802?

These variations are numerous, striking, and fundamental. In many