212 



Proceedings- of the Royal Irish Academy. 



So much it has been necessa^ry to say in order to explain why it is 

 that we cannot be sure of what the reading of B was when we are 

 dependent on Denis's presentation of it. It would be unreasonabh- in 

 the highest degree to expect from Denis what we do not find in TJssher 

 or Ware. The degree of accuracy which is an ideal in the twentieth 

 century was unimaginable in the seventeenth ; and it must not be 

 forgotten that it is chiefly through the Bollandist ^^^^rt Sanctorum \)!\'dX 

 tlie writings of St. Patrick have been known to scholars throughout 

 Europe. Many of the textual emendations that have been adopted 

 in this edition are due to the sagacity of Andreas Denis. 



Even a passing glance at the apparatus criticus reveals the fact 

 that, where A is available, the great majority of the variants from its 

 texts are supported by all the other mss. Accordingly, in these cases, 

 it has been thought best to use the term rell. in place of the full 

 series BCFgEiR or Boll. CF3F4R. Further, the most constant members 

 of this group are CF/\ In fact, neither of these two mss. presents 

 any distinctive readings worth mentioning. In the Confessio C has 

 five omissions by homoioteleuton in §§ 1, 10, 40 his, 60, and some other 

 insignificant variants, most of which are unintentional blunders of the 

 scribe ; and F4* has fewer still. F4 has been corrected throughout in 

 respect of grammatical forms, in so far as was possible to do so by 

 erasure and letters written over erasure. There are no marginal 

 corrections, and hardly any interlinear ones, so that it is impossible 

 to say when the corrections were made. In any case they have no 

 more authority than the emendations of a modern editor. Nevertheless, 

 as being after a fashion ms. evidence, they have all been recorded in 

 tlie foot-notes. 



I have only noticed two cases where C agrees with A against 

 BF3E4, i.e. amor Dei for timor Lei in § 16, and the omission of mihi 

 bef. honor in § 54. The similar agreements of E4 with A are more 

 trifling still, i.e. wide for inde in § 18, and the omission of i7iquit after 

 Dotninus in § 40, sicut Bom, in aeuanguelio ammonet, &c. 



C and E4 agree so closely, not only in genuine variants, but also in 

 blunders and strange spellings of words — they actually have in common 

 an omission by homoioteleuton in Ep. § 13 — that they cannot be con- 

 sidered as independent authorities. As far as the text is concerned, 

 E4 cannot be a copy of C ; both are probably copies of the same 

 exemplar. This, however, does not affect the fact that CF4, which we 

 must treat as one authority, is the most constant member of the rell. 

 group, the common parent of which, although possibly later than A, 

 must have been of considerable antiquity. 



