226 



contain some excellent observations on the forgeries of Macpherson, 

 and some just criticisms on the " post-original GaeHc " of the poems 

 published by the Highland Society of London. 



A poem on the death of the Children of Uisneach, is published 

 in Gillies' Collection, and a much fuller copy in the " Works of the 

 Highland Bards," p. 562. These poems, particularly the latter, are 

 composed of a variety of poems interspersed through the tale of the 

 Death of the Children of Uisneach above mentioned, as may be seen 

 by an inspection of that work. An ancient copy of one of these 

 poems is contained in a copy of the ancient tale, written on vellum, 

 in possession of the Highland Society of Scotland, a fac-simile of 

 which is given in the Report of the Committee of that Societ}^ 

 Appendix, p. 311, Plate HI. No. 4. The two first stanzas only are 

 given in the fac-simile, but the whole nine stanzas, including those 

 two, are given in the Roman letter, with a translation in English, 

 pp. 298 and 299. The manner, however, in which the original is 

 decyphered by the gentleman to whom that task was committed, 

 proves how little qualified even the best Gaelic scholars of Scotland, 

 are to read or translate ancient Gaelic writings. There is a blunder 

 in the very second line, where the word lingantaibh is given instead 

 of hingantaibh. The latter word is the dative case plural of Ingnadh, 

 or longnadh, a wonder. The expositor not understanding the con- 

 traction in the original, mistook the word for lingantaibh, which he 

 translated lakes, supposing, erroneously, that the word linn, a pool of 

 water, formed its dative plural lingantaibh ; but in this he was much 

 mistaken, the word is of one syllable, and could not take three sylla- 

 bles in its oblique cases. The regular dative plural of li7in is linntaibh. 

 There are several other errors in those nine stanzas, but to enter into 

 a particular examination of them, would be foreign to the object of 

 this Essay. The errors here are, however, far exceeded by those in 



