word; therefore we write stick, block, which were originally sticke, blocke. In such words c is now mute."—Johnson. Just as there was a prejudice against i or e ending a word there seems to have been one in the case of c. In the word Frederick there are three modes of spelling: 1. Frederic; 2. Frederik; 3. Frederick. Of these three it is the last only that seems, to an Englishman, natural. The form Frederic seems exceptionable, because the last letter is c, whilst Frederik is objected to because k comes in immediate contact with the short vowel.
Now the reason against c ending a word seems this. From what has been remarked above, c seems, in and of itself, to have no power at all. Whether it shall be sounded as k or as s seems undetermined, except by the nature of the vowel following. If the vowel following be small, c=s, if full, c=k. But c followed by nothing is equivocal and ambiguous. Now c final is c followed by nothing; and therefore c equivocal, ambiguous, indefinite, undetermined. This is the reason why c is never final. Let there be such words as sticke and blocke. Let the k be taken away. The words remain stice, bloce. The k being taken away, there is a danger of calling them stise, blose.
A verbal exception being taken, the statement of Dr. Johnson, that in words like stick and block the c is mute, is objectionable. The mute letter is not so much the c as the k.
"G at the end of a word is always hard, as ring, sing."—Johnson. A verbal exception may be taken here. Ng, is not a combination of the sounds of n+g, but the representation of a simple single sound; so that, as in the case of th and sh, the two letters must be dealt with as a single one.
"G before n is mute, as gnash, sign, foreign."—Johnson. The three words quoted above are not in the same predicament. In words like gnash the g has been silently dropped on the score of euphony (see remarks on k); in sign and foreign the g has not been dropped, but changed. It has taken the allied sound of the semivowel y, and so, with the preceding vowel, constitutes a diphthong.
Before a, o, u (full vowels), g has the sound, as in gay, go, gun: before e, i, y, that of gem, giant.
At the end of a word (that is, followed by nothing at all), or followed by a consonant, it has the same sound that it has before a, o, u—agog, grand. This shows that such is its natural sound. In hedge and oblige the e mute serves to show that the g is to be pronounced as j.
Let there be the word rŏg. Let the vowel be lengthened. Let this lengthening be expressed by the addition of e mute, roge. There is now a risk of the word being called roje. This is avoided by inserting u, as in prorogue. Why, however, is it that the u runs no chance of being pronounced, and the word of being sounded prorogwé? The reason for this lies in three facts. 1. The affinities between the sounds of ga and ka. 2. The fact that qu is merely kw. 3. The fact that in qu, followed by another vowel, as in quoit (pronounced koyt), antique, &c., the u is altogether omitted in pronunciation. In other words, the analogy of qu is extended to gu.
For the varied sounds of gh in plough, tough, enough (enow), through, we must remember that the original sound of gh was a hard guttural, as is at present the case in Scotland, and between g, h, f, v, w, there are frequent interchanges.