Lawi.or — The Cathach of St. Columba. 247 by a margin over 1 6 mm. in width. In that case the height of the pages was not less than 235 mm., and the breadth not less than 155 mm. But the inner measurements of the box which originally contained the manuscript cannot have been greater than those of the present lining of the cumdach, i.e. 220 X 165 mm. The box, therefore, conld not have held the manuscript in its original state ; still less the boards in which it seems to have been bound. Thus we once more reach the conclusion that the Psalter was already mutilated when it was placed in the shrine. Nearly three-quarters of the leaves of the Cathach ajjpear to have been ruled with horizontal lines for the guidance of the scribe,' and probably most of these had also vertical lines separating the margins from the text.* The rules were heavily drawn on the recto pages with a pointed instrument ; and thus for each sunk rule on the recto there was a raised rule on the verso. Consequently, if we allow for a curious tendency of the scribe to write a httle above the rule on the recto, and a little below it on tlie verso, the lines of script on the two sides of a ruled leaf closely corresponded with each other. On the other hand, there are certain leaves in which there can be detected no trace, or only very uncertain traces, of ruling,' and in most of which the lines of writing on verso and recto do not correspond.* It is almost certain that they were unruled. Among these, again, a few have marks such as we see in other leaves of the MS., indicating the ends of horizontal rules :' in spite of this the rules were not drawn.' It is plain, then, that the vellum used by the scribe was of three sorts. Some of the leaves were fully prepared for writing ; on some the process had gone as far as the ticking of the ends of but no trace can be found of the other edge. It is evident that the binder, in separating f. 32 from its conjugate, cut away a wide strip of the vellum. It may be remarked that there were some errors in the numeration mentioned above. Thus ff. 10, 12, 14 were reversed, the numbers being in each case on the verso. The binder has corrected the error in fF. 10, 14 ; but he did not notice it in f . 12, which stUl has its verso to the front. Similarly, ff. 35, 36 were transposed, and still remain so. On ff. 42, 43 the numbers are correctly placed, but the binder has transposed the leaves. There are no numbers on ff. 1-3 ; in the manuscript as bound, they are all reversed, and ff. 2, 3 are, in addition, transposed. Thus ff. 1', 1', 2', 2% 3', 3% 12', 12'-, 35, 36, 42, 43 are in the manuscript numbered respectively 1', 1', 3', 3', 2 2', 12% 12"', 36, 35, 43, 42. These errors are corrected in the text as printed below. 1 Clear traces of the rules are visible on fl". 1-5, 9-11, 13, 19-23, 29-34, 36, 39-45, 49-52, 55, 56. They are more doubtful on ff. 6, 12, 14, 38, 46, 54, 57. In all these leaves, 41 in number, the script on the recto corresponds -nath that on the verso. 2 Visible on ff. 1, 3, 9, 10, 19-23, 29-34, 39-45, 48(?), 49-52, 55. The mutilation of the MS. would cause many of these marginal notes to disappear. ^ None on ff. 7, 8, 16-18, 24-28 ; and apparently none on ff. 15, 37, 47, 48, 53, 58,
- The exceptions are ff. 16, 24, 25, 48, 58.
' ff. 24, 26, 28, 53 (?). Moat evidently so on f , 88, where the points do not tally with the script.
