Lawlok — The Cathach of St. Columba. 287 



C differs from oj3' only in the accidental repetition of dc iic. This would 

 be easily corrected by a scribe coj^ying from C. 



Ps. xcv. «o.>j ecclesiae uocantis ad fidem. 



C, which has the support of P, is probably correct. a|3 omit ad fidtm. 



Ps. xevii. uox ecclesiae ad dominum et cul api'stolos. 



The second ad is omitted by a and the printed text of B. But it is in 2, 

 and was therefore not improbably in /3. 



Ps. ci. uox christi et ecclesiae cum ascendisset ad patrem. 



a and the printed text of B add christus after ascendisset. But C agrees 

 with S, and probably with j5. 



Ps. ciii. uox ecclesiae laiulat dominum opera eius narraiis fideli populo sua. 



This heading is restored with the help of FHMQ. Whether or no the 

 illegible letters have been correctly supplied, it is clear that /3, uox ecclesiae 

 laudantis demti, may have come from it by way of abbreviation ; while C could 

 not have depended on /3. The heading of a, uox ecclesiae ad piopulum 

 suum, might also have sprung from that of C (though not through /j) : but 

 actually it is simply repeated from Ps. cii. 



Let us collect the results to which this discussion leads. In the first 

 place, it leaves the impression that C has a good text in its rubrics. In 

 certain cases we have found it in agreement with a where /3 proljably also 

 agreed with a,^ or where its text cannot be recovered.^ But more striking is the 

 evidence that C has the text of a manuscript from which a and j3 have been 

 derived. Three times the text of a appears to have been derived fi-om that 

 of C, the text of /3 being irrecoverable.'' Once jS is an abridgement of C, while 

 a repeats the rubric of a previous psalni.^ Thrice both o and /3 show signs of 

 derivation from the C text.^ Six times C agrees with [5 against a.' If these 

 cases stood alone, it might be inferred that C has the text of which all the 

 other texts in our group are derivatives. It would be possible even to 

 suggest ihat a and (3 were descendants of the manuscript C itself : a 

 suggestion which would receive confirmation from the curious phenomena 

 of the rubrics of Pss. xli, xlii, xliv. But there are other facts to be 

 taken into account, which negative this hypothesis. There are five places 

 in which C, though differing from a/3, is clearly derived from an exemplar 

 which had their test." In some of these no doubt C differs from his 

 exemplar only by a slip which could easily be set right by a subsequent 



' See above, p. 273. ^Pss. slv, xlviii, Isssis, xc- 



^Pss. xliii, liv, Ixvii. ■'Pss. xxxvi, sxxvii, xli. 



^Ps. ciii. ^Pss. xlii, xliv, xcv. 



' Pss. xxxvui, xlvi, 1, Ixxviii, xcvii, ci. 

 ^ Pss. Ivii, Ixxxii, Ixxxiii, Ixxxvi, scui. 



