Some of your correspondents have taken up the not unnatural idea, that
the last syllable of the word "Armagh" is identical with the Celtic
word _magh_, a plain. But there are two objections to this. In the
first place, the name is never spelt in Irish _Armagh_, nor even
_Ardmagh_, but always ARDMACHA. _Ardmagh_ or _Armagh_ is only the
anglicised spelling, adapted to English tongues and ears. It is
therefore clearly absurd to take this corrupt form of the word as
our _datum_, in the attempt to search for its etymology. Secondly,
the Irish names of places which are derived from, or compounded of,
_magh_, a plain, are always anglicised, _moy, moi, mow_, or _mo_,
to represent the pronunciation: as Fermoy, Athmoy, Knockmoy, Moira,
Moyagher, Moyaliffe (or Me-aliffe, as it is now commonly spelt),
Moville, Moyarta, and thousands of other cases. And those who are
acquainted with the Irish language will at once tell, by the ear, that
_Armagh_, as the word is pronounced by the native peasantry, even by
those who have lost that language (as most of them in that district
now have), could not be a compound of _magh_, a plain.
The work of M. Bullet, quoted by your correspondent "HERMES," is full
of ignorant blunders similar to that which he commits, when he tells
us that Armagh in compounded of "_Ar_, article, and _mag_, ville."
The article, in Irish, is _An_, not _ar_; and _mag_ does not signify
a town. He adopts, your readers will perceive, the modern English
spelling, which could not lead to a correct result, even if M. Bullet
had been acquainted with the Celtic languages. The same remark applies
to the explanation given by the author of _Circles of Gomer_. _Ard_,
not _Ar_, is the word to be explained; and therefore, even though _Ar_
and _Ararat_ meant, as he tells us, "earth, country, or upon and on
the earth," this would throw no light on the etymology of ARD_macha_.
"HIBERNICUS" (No. 14. p. 217.) is partly right and partly wrong; he
adopts the anglicised spelling of the second syllable, although he
seems aware that the first syllable ought to be _Ard_; and he admits
also that this word is a substantive, signifying a _height_, not the
adjective _high_. "A high plain," in Irish, would be, not Ardmagh,
or Ardmoy (as it would have been anglicised), but _Magh-ard_ (Anglice
_Moyard_). Great light will be thrown on the whole subject of the
etymology of Irish typographical names, when the Index to my friend
Mr. O'Donovan's edition of the _Annals of the Four Masters_ makes its
appearance.
I may add too, in conclusion, that Camden is wrong in suggesting that
_Armach_ (as he spells it, retaining, curiously enough, the correct
etymology of the last syllable) is identical with _Dearmach_ (where
the last syllable ought to be _magh_). This latter place is the
well-known Durrow, in the county Westmeath; and its name, in Irish,
is _Duir-magh_, which is really a compound from _magh_, a plain. Bede
tells us, that the word signified, in the Scottish language, _Campus
roborum_ (see Bede, _Hist. Eccl._ lib. iii. c. 4.); but Adamson (_Vit.
Columbæ_, c. 39.) more correctly translates it, "monasterium _Roboreti
Campi_." It is not likely that such authorities could confound Durrow,
in Westmeath, with the ecclesiastical metropolis of Ireland, and
patriarchal see of St. Patrick.
Whoever the Mach or Macha was from whom Ardmacha has its name
(whether the queen called Macha-mong-ruadh, whose reign is assigned
by O'Flaherty to A.M. 3603, or the older Macha, who is said to be the
wife of Nemedius), it should be borne in mind, that the word whose
etymology is required is ARDMACHA[16], and not _Armagh._ What would
be thought of the critic who would now attempt to investigate the
etymology of the English word _bishop_, by dividing it into two
syllables, and seeking analogies in sound for each syllable.
I have ventured to go at greater length into this matter than its
importance may seem to warrant, because it illustrates so clearly a
very general error, from which Celtic literature has deeply suffered,
of inventing fanciful etymologies adapted to the modern English
spellings, instead of the original Celtic forms of names; and this
error, as the question before us proves, is as old as Camden's time,
and older.