us. And let macLonan's Songbook be given to me, that I may understand the
sense of the poems that are in it. _Et vale in Christo._"[9]

It would seem from another note in the manuscript[10] that the Book of
Leinster afterwards belonged to some admirer of King Dermod, for he wrote:
"O Mary! Great was the deed that was done in Ireland this day, the kalends
of August (1166)--Dermod, son of Donnoch macMurrogh, King of Leinster and
of the (Dublin) Danes to be banished by the men of Ireland over the sea
eastwards. Woe, woe is me, O Lord, what shall I do!"[11]

My reason for founding the translation on the LL. version, in spite of the
fact that its composition is posterior by half a century to that of LU.,
was not merely out of respect for the injunction of the scribe of the _ne
varietur_ and to merit his blessing (page 369), but also because LL.'s is
the oldest _complete_ version of the Táin extant. Though as a rule (and as
is easily discernible from a comparison of LU. and LL.), the shorter,
terser and cruder the form of a tale is, the more primitive it is, yet it
is not always the oldest preserved form of a work that represents the
most ancient form of the story. Indeed, it is not at all improbable
that LL. contains elements which represent a tradition antedating the
composition of LU. At all events, LL. has these strong points in its
favour, that, of all the versions, it is the most uniform and consistent,
the most artistically arranged, the one with most colour and imagination,
and the one which lends itself most readily to translation, both in itself
and because of the convenient Irish text provided by Professor Windisch's
edition. In order to present the Táin in its completest form, however, I
have adopted the novel plan of incorporating in the LL. account the
translations of what are known as conflate readings. These, as a rule, I
have taken from no manuscript that does not demonstrably go back to a
twelfth or earlier century redaction. Some of these additions consist of