Lawlok — The Cathach of St. Columba. 249 



superfluous letter, which was afterwards deleted. For example, at xxxiii. 21, 

 the second n of conterentur has been erased, at Ixvii. 23 apparently 

 the first s of bassan, at Ixviii. 8 the t of operuit, at Ixx. 20 the first s 

 of quantass, at Ixxxii. 19 the first ^ of coggnoscant, at Ixxxiii. 6 the 

 second s of disspossuit, at civ. 35 the second in of commedit. May we 

 suppose then that in all cases the erased letter was one that had been intro- 

 duced by mere Icqjsus calami, and was subsequently removed ? If so, we 

 must assume that the insertion of letters was the most frequent error of the 

 scribe ; and yet one that was almost always detected and set right : for I 

 have noted only eighteen such mistakes which have not been corrected.' On 

 the other hand, omission of letters, to which he was also prone, has been 

 detected by him, as we have seen, but rarely. But in two of the instances 

 just mentioned the effect of the erasure is not merely the removal of a super- 

 fluous letter, but the substitution of one part of a verb for another. In 

 xxxiii. 21 the plural is altered to the singular, and in Ixviii. 8 the third person 

 to the first. Another case of the same kind is instructive. In Ixxxvii. 16 

 the words a nmientus occur at the end of a line, the letter s being erased. 

 Above the line, after the manner of Irish scribes, is written / ta meet. Thus 

 iimentus is transformed into iuuentute. We may affirm, with a probability 

 not far removed from certainty, that the scribe perceived his error imme- 

 diately after he liad written iimentus, and erased the s before he penned the 

 conclusion of the clause. Now a large number of the erasures which we are 

 considering are patient of a similar explanation.- In very many cases the 

 erasure immediately precedes the termination of a verb or substantive; and 

 we may well believe that the scribe had caught himself in the act of writing 

 a wrong termination, and made the necessary correction immediately. Other 

 examples . of the scribe's habit of correcting himself as he went along are 

 forthcoming. In some instances between two words of the text there is an 

 erased letter which can scarcely have been written as part of either of 

 them. Thus, xxx. 5: mcditimn /-^autem:. Here the scribe may have 

 begun to write autem, omitting the obelus. Similarly he would seem in 

 Ixi. 10 to have penned the first letter of some other word instead of uani, the 

 first letter of a substitute for a at Ixvii. 9, and of a substitute for uirtns at 

 Ixx. 9. At Ixviii. 33 he may, perhaps, have written animae ; but the error 



' xxxii. 16 ; xxxiv. 8; xxxvii. 12 ; xliii. 19, 23; xlviii. 16 ; liv. 10 ; Ixii. 1 ; Ixvii. J 3 ; 

 Ixxv. 6 ; Ixxviii. 8 ; Ixxx. 17 ; Ixxxii. 6 ; Ixxxvii. 16 ; Ixxxviii. 10 ; xciii. 2 ; ci. 29 ; civ. 

 30. 



^M.g., xxx. 23 ; xxxiii. 9, 21 ; xxxvii. 12 ; xlix. 21 ; Iviii. 16 ; Ix. 3 ; Ixiii. 6 ; Ixviii. 

 8, 10 ; Ixxi. 9 ; Ixxiii. 2 ; Ixxv. 12 {ditu, corr. from do ?) ; Ixxvii. 3, 32, 55 ; Ixxviii. 11 ; 

 Ixxxviii. 11 ; xcv. 7 ; cii. 13 ; civ. 40. 



B.I. A. PEO ., VOL. XXXni., SECT, o. [37] 



