300 Proceedings of the Royal Irish Academy. 
can be said to be distinguishable, these followed, after a short lacuna, 
by N. Directly following the N comes a well-marked M, a lacuna 
equal to about four points, a Q, and another lacuna, suggesting very 
cogently the lost Q and I of an original Magi, written Megi, or it 
may be Deqq, or Migg. Then comes the name of the parent, a geni- 
tive in a, and we note the seemingly feminine form with increased 
interest, observing that the word begins with double L, followed by a 
lacuna long enough to hold the vowel points, which would yield z @ 
before the existing N A at the end, and so give /liana. Here, then, 
would be Magqgq lliana in conjunction with what appear the elements 
of the name of Merlin Map leian. 
Applying our attention now to the principal name, we look in vain 
for any trace of the missing M immediately before the E R L; but, 
recalling the fact that the triradial groups with their string of vowels 
of the eastern arris are preceded by the compound character contaming 
that letter, and that this is the initial character of the entire legend, 
we may be reminded of something similar which led me, so far back 
as 1870, to speculate on the probable existence in Ogham legends 
of what I ventured to designate as the ‘‘ dispartition of proper names,” 
on which analogy it might be allowable to accept the initial M of 
the opposite angle, whether compounded or distinct, as the desiderated 
initial of the name, which would thus assume the nearly complete 
form, MERL N. 
We might accept the uninflected N before Maqias the last character 
of the name. But the interspace is too long for a single 7, and must 
have held at least twice as many points and digits. It might have 
held both the ¢ and ng necessary to complete Merling, which would 
account for the uninflected N before Maqi. But if this last letter be 
not part of the name, some vowel must have preceded it to give it an 
independent articulate force. What must we assume this to be? The 
question receives a solution agreeable to the hypothesis which has 
conducted us so far, in the prefix anmap, as we have seen it above 
associated with the name of Merlin. This concludes the reasoning on 
which it is submitted, that if the entire inscription, omitting the 
interjected symbols, and their string of vowels, were spread before us, 
as it was originally sculptured, it would present this appearance :— 
M [ |] ERLING ANMEQQ LLIANA 
Meriin m‘s [born] son of the Nun. 
T do not suggest that this was the sepulchral monument of Merlin, 
supposing such a person ever to have existed. Its position seems 
rather that of a termon pillar, looking to the neighbouring ecclesias- 
tical precinct of Eglys Nunydd, distant about three hundred yards. 
‘The old buildings at Eglys Nunydd have been partly incorporated into 
a modern residence, but the antiquity of the site is evinced by a 
sculptured stone in the grounds, which bears the outlines of a Greek 
