Discusses numerals in Old Breton glosses in MS Angers, Bibliothèque municipale 477: 1. ‘thirty’; 2. ‘thirtyfold’; 3. ‘sixty’; some comparisons with Old Irish forms.
Derives the feminine forms of ‘3’ and ‘4’ from a Proto-Indo-European ablauting paradigm with amphikinetic accentuation containing the feminine formans *-s(o)r-.
Argues that an Indo-European syntactical rule underlies the Old Irish (and also Welsh) omision of the ordinal ‘first’ in complex numeral phrases, for which an Albanian parallel is found, and that in both these languages this deletion is replaced by an new form for ‘first’ based on the cardinal.
[1.] general usage; [2.] Use of plural forms (units of measurement, set phrases); [3.] Irregularities of mutation. Also on the lenition of déag and fichead.
Rejects (with W. Cowgill 1957; see BILL 3082) the view that the Milan disyllabic hapax teüir reflects a morphological archaism of Indo-European (i.e. an ablauting stem containing a feminine suffix *-sor-; cf. K. McCone, in Ériu 44 (1993), pp. 53-73), and argues that the Celtic forms are best understood as continuing uniform proto-Celtic stems *tisr- and *kwtesr-.
Appendix A: On the masculine and neuter forms of ‘3’ and ‘4’.
Appendix B: PC *-Vsr- in Irish and British.
‘Man / warrior of the pair’: diad related to dïas ‘two persons’. Also suggests i ndiaid may represent ‘in (its) pairing’ rather than ‘in(to) (its) end’ (cf. i ndiad).
ad R. Hertz, in Lexis 4 (1955) 66-69. Derives dëec from *dechǣg (< PC *dekank < *deḱm-kwe ‘and ten’) with dissimilatory loss of *-ch- (or *-k-) before *-g-.