426 
Proceedings of the Royal Irish Academy. 
On the Relative Ending -AS, [-^AS]. 
The doctrine of the Eelative clause, which in the early Irish 
prose is fairly consistent and intelligible, has been so obscured 
by groundless theorising and illegitimate deductions, that a suc- 
cinct statement of the excellent practice of the early writers can 
hardly fail to be of service. 
In modern Irish there is no word or particle of any kind 
whatever to express the ' relative ' connexion of two clauses, 
unless in cases where the construction stands in need of a pre- 
position to govern such (relative-) word : the relative-connexion 
is expressed simply by direct sequence, and that whatever be 
the tense. 
The precise relation, however, of antecedent to relative 
is not determined in this mere juxtaposition : cf. Latin homo 
[_gue?n] cimavit ; and we have therefore to consider what difficul- 
ties might arise, and how these are obviated in Irish. In such 
sentences as ' the man whom he struck,' it is evident that even 
if the relative ' whom ' were omitted, the relation of the clauses 
is clearly determined by the pronoun he. Again, in ' the sen- 
tence * the man whom John struck,' the relation of the clauses, 
even in the absence of the relative, is manifested by the position 
of the words ' John ' and * struck,' and in fact we frequently 
omit this acc. relative, when the subject of the relative clause is 
named, e. gr. ^ the man he struck,' ' this is the man John struck,' 
where the sense is quite intelligible. Now, the point to be dis- 
tinctly grasped, in Irish syntax of relatives, is that there is no 
WORD CORRESPONDING TO THE ENGLISH ^ WHO ' OR * WHOM,' UulcSS 
where the * whom ' is governed by o. preposition. But with the ap- 
pearance of a preposition, though the mere sequence is often quite 
sufficient, — for even in English, we say, ' the subject he talked of/ 
* the man he met with,' and in modern Irish, where the composite 
prepositional pronoun is at hand, these sentences, xsn nix), tdo u|iacc 
-pe £>.n i:e^|\ "oo cx^i^x^'o m\\ [orteip], are perhaps even com- 
moner than in English, — yet neither in English nor in Irish is 
such construction felt to be correct ; and we ivrite, * the subject on 
