that one end of a handle might have been secured to it. When fitted into it, the horizontal sunk plane would have steadied and given firmness to such a handle. The whole arrangement suggests that the mount formed a portion of a handled tankard. The Late Celtic Trawsfynydd tankard is the outstanding example of a vessel of this character. This tankard is composed of wooden staves covered with sheet bronze. The ornamental metal work is confined to a knob, with serrated edge in the centre of the bottom of the tankard outside, and to its wonderfully beautiful and elaborate handle. It should be noted that this handle is secured by four rivets with round knobs on the outside, arranged to form part of the decoration, and going right through the bronze plating and the staves. The arrangement of rivets in the Braich y Dinas specimen would appear to have been on the same principle, but, in this case, there would have been only two rivets, one securing the top, the other the bottom, of the handle. In the Trawsfynydd tankard there is no elaborate ornamental band below the rim. The rivets in the Braich y Dinas bronze, with their elaborately orna- mental heads, forming part of the decoration, are arranged to secure the staves of which, probably, the vessel itself was constructed. With reference to other finds from Braich y Dinas, published in former reports, much of the datable pottery is of the late first or the early second century. Other fragments are of third-fourth century type. The datable coins are of Nerva, Trajan, and Hadrian. I have to thank Mr. W. Aspden for the photograph of the bronze ornament. One or two other finds, but of no great archaeological interest, have lately been brought to light during quarrying operations. One is a fragment of a saddle quern of Millstone grit, with one face slightly concave curved the other is a spindle-whorl of fine-grained sandstone. Both these were found together just outside the outer defence wall, shown in solid black, at the north end of the fortified hill-village, beyond Hut Z2 on my plan of 1923, at an altitude of about 1,380 1,400 feet. H. HAROLD HUGHES. NOTES ON NAMES IN THE Mabinogi. — Pwyll. Lady Guest, seeking to identify Pwyll with some historic person, refers to pedigrees, where he is said to be the son of Argoel or Aircol Law Hir, son of Pyr y Dwyrain {Everyman, 287). This is interesting, for it looks like a garbled version of an attempt to equate Pwyll with VortiPorius, or more rightly Voteporix, whose tombstone was at Castell Dwyran, son of Aircol, i.e. Agricola, king of Dyfed. Castell Dwyran, I may say, is locally pronounced Casdiran. Mr. Timothy Lewis {Mabinogi Cymru, 18, 47-8), strongly reacting to the findings of Sir John Rhys and his school, explains Pwyll as from pugil, champion fighter, and would identify Pwyll of Dyfed with a Scandinavian sea-rover from the Faroe Islands, whose name may have been Sigmund, perhaps the Symon of The Rath in Rudbaxton (in Welsh Rhyddbac), near which was a St. Leonard's Chapel, Capella s. Leonardi de Castro Symonis (Fenton, 1903 ed., p. 196) thereabouts also, I may add, was Simonstown (Rep. of MSS. in Welsh, i, 917). Mr. Lewis rejects the identifi- cation of Pwyll of Annwn with Pwyll of Hell. Rhys and his school found it easy to believe that the Welsh journeyed back and fore to Hell at will. Os mynnai Sais, etc., if an Englishman or Germans or others wanted to go to Hell, they had to