Page:Welsh Medieval Law.djvu/436

 indeed to have been used for that remnant of independent or semi-independent territory which was still left in the hands of the princely house of Dinevwr, but Deheubarth was never rightly the name of a definite patria or gwlad. The only other reference to Deheubarth in our present text is in the opening preface, where it is attended with considerable difficulties, for mention is made of its sixty-four cantrevs, an obviously impossible number. Indeed, the whole of this passage, wherein Howel's dominions are enumerated, is full of difficulties. The passage, which it will be convenient to quote here, is virtually the same in all the texts, with the exception of Z (Peniarth MS. 2596 of the sixteenth century). It is as follows:— ' petwar cantref a thrugein Deheubarth, a deunaw cantref Gwyned, a thrugein tref tra Chyrchell, a thrugeint tref Buellt.' According to Aneurin Owen, the MSS. U, Y, and Z place yn before Deheubarth, whilst Z changes the first a thrugein into arhugain, thus reducing the sixty-four cantrevs of Deheubarth into twenty-four, a facile alteration made by a late writer, which hardly diminishes the difficulty. We may therefore safely treat the passage as meaning 'sixty-four cantrevs of [or in] Deheubarth, and eighteen cantrevs of Gwynedd, and sixty trevs beyond the Cyrchell, and sixty trevs of Buallt '. The first point to notice is that Powys proper is clearly omitted and also the patria of Rhwng Gwy a Havren with the exception of tra Chyrchell, i.e. Gwrtheyrnion, which here, as since the days of Pasgen ab Gwrtheyrn in the fifth century, went with Buallt. Let us note further that tra Chyrchell, beyond the Cyrchell, as referring to Gwrtheyrnion, must have been used by a person speaking and writing east or north-east of the brook Cyrchell, that is to say, by a person living in the patria of Rhwng Gwy a Havren or possibly in Powys proper ; at any rate within that part of Wales which the writer carefully excludes as belonging to Howel's dominions. The fact that Buallt is mentioned after 'tra Chyrchell ' strengthens the argument. Our present author therefore (possibly Cyvnerth ab Morgeneu) appears to be outside the Deheubarthwyr or Dextrales, and it may be that he is one of the Powyssi. The next point is the number of cantrevs given to Deheubarth and to Gwynedd, sixty-four to the former and eighteen to the latter. As there were never sixty-four cantrevs in the whole of Wales, and as the highest number given to Gwynedd in the old lists is eleven, it is clear that there must be some error in the text. If we assume for a moment that the original of this passage in our preface was in Latin, the word