1078 Wisconsin Academy of Sciences, Arts, and Letters. 
velopment of E to I seems plausible, when we remember forms 
like meta, eta, e’ reduced from meita, etc. (see §85) and not 
proceeding to mita, etc. The form el for il (see §37, note 7), 
:£cmnd quite commonly in the old Florentine writers, might 
then be explained in either of two or three ways,—it might 
represent the last moment before passing to I, or it might have 
been perverted or delayed in its development by analogy of the 
dissyllabic forms dello, etc. 1 
8. Concerning the development of O O XJ, Professor Grand- 
gent, in a personal note to the writer, states: “I think we must 
assume a popular tendency to u and a more cultivated to o, 
under all conditions, and perhaps all the differences not due 
to dialect, nor to bookish influence, may be explained in this 
way. However, the examples seem to show that 1 (and pos¬ 
sibly unaccented i and palatal) favored u in Florence. As 
to fj> u, the development is evidently not uniform, although 
the majority of examples show u. One might cite P. Haberl, 
in ZPPh XXXIV 147, but I don’t think he is right.” Pa- 
rodi, also, in Giorn. Stor. X 182-3, says: “Si noti che il fioren- 
tino dice uliva, ufficio, non oliva, officio, e oosi mulino, fru- 
mento, ed ora anche ugni eosa, urecchio, pumpiere, scudella, 
ecc.” The material is certainly very contradictory and per¬ 
haps my zeal to obtain as formal rules of development as pos¬ 
sible carried me further than was wise. 
IY (long) 
1. Classic Latin I Y (long) > Vulgar Latin I>Italian I. 1 
§7. Note 1. See §20, note 6. Parodi, 1. c., pp. 187-8, says: “Xo os- 
servero che e importante il veder gih qui stabiliti il ed i; che nel 
fiorentino questa e regolare, nonostante cio che dice il D’Ovidio (AG 
IX, 100), pe'rche igli potrebb’esser bene in esso la forma originaria; 
che el inoltre, contro cio che cornunemente si crede, e frequentissimo 
nei testi fiore'ntini del secolo XIV e XV, dimodoche senza dubbio 
questa forma, che e quella che normalmente si attenderebbe da “ille,” 
fu anche nei secoli anteriori ben viva. Tuttavia che i sia rifatto su 
il, ch’e poi un prodotto della tendenza italiana ad i atono, secondo la 
teoria del Diez, perfezionata dal Caix e dal D’Ovidio, vien confer- 
mato dal prevalere d’il, i (e di in-, anziche en-) anche nel piu antico 
senese.” 
§1. Note 1. It was thought best not to group together the develop¬ 
ment of i and i. As a matter of fact, although the results in Flor¬ 
ence are regularly the same, the principal difference lies in the ex¬ 
ceptions; the double forms from I are rare, whereas from i they are 
very numerous,—see under §§3, seq. 
