Palamedes's invention. And it may be Ch. 3. obferved, that this way of marking the afpiration, fhews the nature of it better than Palamedes's characters: for they have probably led the grammarians into what I hold to be a mistake, that the aspiration belongs to confonants; whereas it belongs, as I have faid, primarily to vowels, and only confequentially to the confonants, by their being joined to afpirated vowels. Thus when I write Cethegus in Latin, it is evident that it is the vowele that is afpirated; whereas, when I use Palamedes's character, and write Kɛ☺nyos, it may be thought, and is commonly thought, that the afpiration belongs principally to the confonant t. And as to Simonides's marks for the long and long, if he had carried the invention farther, and devised marks for all the long vowels, it might have been fo far useful, that it would have faved writing; but as he did not carry it fo far, he had much better have let it alone altogether, and then it is likely the old way of writing would have continued, of doubling the character when the vowel is long, of VOL. II. L1 writing,

દ

Ch. 3. writing, for example, aa, when the a was long *.

E

But though these additions, made by the Greeks to the Egyptian alphabet, were neither neceffary nor ufeful; fome corrections which they made upon their own orthography, were certainly a great improvement. For the letter, among them, ftood for three different founds, the short , the long ; and the diphthong e, which was the name of the letter. The long they expressed, as I have juft now faid, by doubling the character; but it was certainly very blundering to make the letter ftand for the name they happened to give it. They might as well have made a ftand for the found apa, or 6 for 67a. In like manner the letter • stood not only for both long and fhort, (which ambiguity was removed by writing it double), but alfo for the diphthong, probably for the fame reason that stood for e, namely, because was the name they gave the let

ter.

*This was the antient practice among the Latins, as appears from Quinctilian; and probably alfo among the Greeks. See what I fay further of this fubject, when I come to treat of accents.

Several

Ch. 3;

Several other obfervations might be Ch. made upon the Greek alphabet, but enough has been faid of the alphabet of a particular language and I fhall conclude this part of the analysis of language, after having made a few observations upon the Roman and English alphabet.

The Roman alphabet was, as I have fhewn elsewhere, the antient Greek alphabet, probably more antient than that which Cadmus brought into Greece; and

as I have faid, it was no worse for wanting the additional letters invented by Palamedes and Simonides.

was a dialect of the Eolic,

But as the Latin

and as the olic ufed very much the found of the digamma, which resembled the found of our w, the Romans had the fame found, but did not use the character, making the letter u and v (for both forms were used indifcriminately) ftand both for the vowel and the digamma; for as to their letter f, it neither expreffed the Greek, nor the Eolic digamma, but a found different from either, and a very unpleasant one, as appears from the paffage above quoted from Quinctilian, lib. 12. cap. 10. To fupply this defect in the L12 Latin

Latin alphabet, Claudius, the Emperor, introduced the ufe of the Æolic digamma, marked like a Roman F reverfed, which is ftill to be seen in fome antient infcriptions, but went out of use after his death *.

yet

As to our English alphabet, it is certainly very faulty. For the first letter has three founds: first, the common found of a; then the found of the diphthong au; and lastly, the found of the Greek; and there is but one character to exprefs all the three. Then the i fupplies the double office, expreffing both the genuine found of that letter, and of the diphthong ai. E, in like manner, ftands both for its own found, and the found of i; and u is fometimes the diphthong eu, and fometimes the plain vowel, or rather the diph

*See what Mr Fofter has collected upon the fubje& of the digamma, pag. 122. of his Effay, to which may be added the paffage I quoted above from Quinctilian, lib. 12. cap. 10. ; and alfo what Mr Fofter has observed from Mr Dawes, concerning the effect of this digamma, in making the preceding vowel long, as in the perfect tenfes of the third and fourth Latin conjugation; for in the preterite cupii and audii, the firft is fhort, but by inferting the digamma betwixt them, it becomes long, as in cupivi and audivi.

thong

thong; for, as I obferved, we have not Ch. 3. in English the genuine found of that vowel. Then, with respect to confonants, the c is an ambiguous character; for it is fometimes founded hard as the k, and fometimes foft as the /; and the t is of ten founded as fh. And in the combinations of confonants in fyllables, we do not always give them the fame found; for the th in thing is a much stronger afpirate than in then or though,

Of the antient accents.-That they were real notes of mufic, diftinct from the quantity of the fyllable.-What accent in English

is?

I

Come now to the analysis of the second Ch. 4. part of the matter of language, of which I proposed to treat, viz. the Profody. And here I am to speak of a thing fo little understood in modern times, that fome even deny the existence of it; I mean, the melody of language, as the antients called