e in which Agricola is said to have encouraged
the children of British chieftains to acquire a taste for liberal
studies, and to have succeeded so much by judicious commendation of
their abilities, ut qui modo linguam Romanam abnuebant, eloquentiam
concupiscerent. (c. 21.) This, it is sufficiently obvious, is very
different from the national adoption of Latin as a mother tongue.

[483] t. vii. preface.

[484] It appears, by a passage quoted from the digest by M. Bonamy, Mem.
de l'Acad. des Inscriptions, t. xxiv. p. 589, that Celtic was spoken in
Gaul, or at least parts of it, as well as Punic in Africa.

[485] Atque eadem illa litera, quoties ultima est, et vocalem verbi
sequentis ita contingit, ut in eam transire possit, etiam si scribitur,
tamen parum exprimitur, ut _Multum ille_, et _Quantum erat_: adeo ut
pene cujusdam novae literae sonum reddat. Neque enim eximitur, sed
obscuratur, et tantum aliqua inter duos vocales velut nota est, ne ipsae
coeant. Quintilian, Institut. 1. ix. c. 4, p. 585, edit. Capperonier.

[486] The following passage of Quintilian is an evidence both of the
omission of harsh or superfluous letters by the best speakers, and of
the corrupt abbreviations usual with the worst. Dilucida vero erit
pronunciatio primum, si verba tota exegerit, quorum pars devorari, pars
destitui solet, plerisque extremas syllabas non proferentibus, dum
priorum sono indulgent. Ut est autem necessaria verborum explanatio, ita
omnes computare et velut adnumerare literas, molestum et odiosum.--Nam
et vocales frequentissime coeunt, et consonantium quaedam insequente
vocali dissimulantur; utriusque exemplum posuimus; Multum ille et
terris. Vitatur etiam duriorum inter se congressus, unde _pellexit_ et
_collegit_, et quae alio loco dicta sunt. 1. ii. c. 3, p. 696.

[487] Tiraboschi (Storia dell. Lett. Ital. t. iii. preface, p. v.)
imputes this paradox to Bembo and Quadrio; but I can hardly believe that
either of them could maintain it in a literal sense.

[488] M. Bonamy, in an essay printed in Mem. de l'Academie des
Inscriptions, t. xxiv., has produced several proofs of this from the
classical writers on agriculture and other arts, though some of his
instances are not in point, as any schoolboy would have told him. This
essay, which by some accident had escaped my notice till I had nearly
finished the observations in my text, contains, I think, the best view
that I have seen of the process of transition by which Latin