The experience of every one proves how different is the impression which the sound of a foreign language makes upon the ear, when spoken by another, from what it makes upon the eye when read even by one who is perfectly acquainted with the theory of pronunciation. Until the ear is habituated, it is easier for an Englishman to speak French than to understand it when spoken. If we consider attentively the manner in which we speak our own language, it is astonishing how many letters and even syllables are slurred over and omitted: the accented syllable is strongly and firmly enunciated, the rest, especially in long words, are left to take care of themselves, and the experience of the hearer and his acquaintance with the language find no difficulty in supplying the deficiency. This is universally the case, except in careful and deliberate reading, and in measured and stately declamation. With regard to the classical languages, the foregoing observations hold good; in a slighter degree, indeed, with respect to the Greek, for the delicacy of their ear, their attention to accent and quantity, not only in poetry but in oratory, and even in conversation, caused them to give greater effect to every syllable, and especially to the vowel sounds. But even in Greek poetry elision sometimes prevents the disagreeable effect of a hiatus, and in the transition from the one dialect to the other, the numerous vowels of the Ionic assume the contracted form of the Attic. The resemblance between the practice of the Romans and that of modern nations is very remarkable; with them the mark of good taste was ease—the absence of effort, pedantry, and affectation. As they principally admired facility in versification, so they sought it in pronunciation likewise. To speak with mouthing (hiulce,) with a broad accent (late, vaste,) was to speak like a clown and not like a gentleman (rustice et inurbaniter.) Cicero[170] admired the soft, gentle, equable tones of the female voice, and considered the pronunciation of the eloquent and cultivated Lælia as the model of purity and perfection: he thought that she spoke as Plautus or Nævius might have spoken. Again, he speaks of the habit which Cotta had of omitting the iota; pronouncing, for example, dominus, dom’nus, as a prevalent fashion; and although he says,[171] that such an _obscuration_ argues negligence, he, on the other hand, applies to the opposite fault a term (putidius) which implies the most offensive affectation. From these observations, we must expect to find that Latin as it was pronounced was very different from Latin as it is written; that this difference consisted in abbreviation either by the omission of sounds altogether, or by contraction of two sounds into one; and that these processes would take place especially in those syllables which in poetry are not marked by the ictus or beat, or in common conversation by the stress or emphasis. Even in the more