In view of this possibility of confusion of forms one must check the evidence of the Annals, step by step, by the various Old and Middle Irish documents which can be approximately dated.
Even where it is not possible to work out from the text an approximately exact date for a certain change, as in the case of -o and -a in the genitive singular of u- and i-stems, I hope the material collected and arranged will still be of value as showing what was actually written and what was possible in the orthography of the Annals, and other Old Irish documents at a particular date.
When all the various changes here discussed are taken into account it will be possible to decide with more or less certainty the date of Old or Early Middle Irish texts and to estimate with greater accuracy their relative value or authenticity.
I have in the course of the work compared forms from various sources, but it is for the present impossible to note all the instances which occur of an interesting Old Irish form. A careful consideration of the Old Irish texts in LU such as Compert Conculaind and Fled Bricrenn, and Longes mac n-Usnig, Táin Bó Fráich, etc., in LL. with a view to deciding how far the text has been tampered with by the compiler, would be sure to yield valuable results.
§ 5. I give, throughout the dates of the Annals themselves, which are antedated by one year from 486 till 1013 (cf. MacCarthy, Introduction to the Annals, xcvi. seq.) and not the corrected date. The years 891 to 1012 are corrected in the MS. (H 1. 8) and 1013 is a blank, after which they bear the correct date.
