been without it, and (so being without it) have sounded the combination c+e according to its natural power. We have it, it is true; but we may have evolved it during the interval between their time and ours. Now, though the specific question as to the actual A. S. usage stands over, the more general phonetic problem as to the development, evolution, or genesis of the sound of tsh out of that of k +-y, will be taken up; the best preliminary to which is the reconsideration of the phenomena presented by t+y. The combination t+y became t+sh; in other words, tya became tsha. Assuming the same instability for k+y, what ought that combination to become? Answer; k+ sh. In other words, kya should become ksha. But this is not the case. If it were so, choose would be sounded kshoose, which it is not. Its true sound (as aforesaid) is that of tshoose. Nevertheless, k+y is an unstable combination, and k+sh is the combination into which it is expected to pass. This is a real sound. In certain languages it occurs as definitely and as prominently as the tsh (ch) in English. In such cases its nature is often explained to the Englishman by means of the word election; from which if we subtract the first three letters the sound of -kshon remains (ele-ction). Whether this be the best illustration of the compound sibilant whose elements are k+sh, is doubtful. I give it, however, in order to show the extent to which the combination itself has been recognised. Now this ksh is, itself, an unstable combination; an unstable combination with a tendency to become tsh. No languages show this better than the Norse or Scandinavian. c, and je of the Anglo-Saxons, so that The A. S. orthography would be, In Norse. Therein k The combinations, then, kj and ce are equivalent to each other, i. e. each element in each has the same power-the words in the second column being not absolutely Norse words, but A. S. words as they would appear if written according to the Norse orthography. If we now ask how this kj is sounded in Scandinavia, we get a most instructive answer. In Denmark it either retains its natural power of k + y, or else drops the sound of y and becomes a simple k. In Norway it takes the sound of ksh. In Sweden that of tsh. The Norwegian form I believe to be transitional. We may now go back again to t+y. Doing so, we shall raise a new question. Mutatis mutandis, the history of t+y was also the history of d+y; so that when one became tsh the other became dzh. Assume that gy comports itself in the same way toward k+y, and, mutatis mutandis, the phenomena we have just investigated repeat themselves. Hence, when kya becomes ksha and tsha, gya becomes gzha and dzha. If so The A. S. ge should coincide with the English j; Yet such is not the case. The Norwegian and Swedish gj are often sounded alike, and that, not as either gzh, or dzh, but as the semi-vowel y-gjöre in Norwegian, göra in Swedish, being pronounced yöre and yöra. and so on with the majority of examples of this combination. This departure on the part of g+y from the analogy of d+y, presenting itself as it does (to go no further) in three languages, is remarkable. It can, however, be explained; the close affinity between g and y (ga and ya) being the reason of it. It is this affinity which diverts (so to say) the line of change from the compound sibilants to the semi-vowel. The evolution of the sound of tsh (for an evolved rather than an original sound it seems to have been) has now been explained; as has, also, the reason why gy has not undergone, mutatis mutandis, the same series of changes as ky, and it may be added that our inquiry, as far as it has hitherto gone, has been to the effect that the sound of dzh should have no existence in English, except so far as it is either an original element of the A. S. phonesis, or a result of the combination d+y (ver-dure, ver-dzhoor). At any rate, it has not grown out of the process which gives us our tsh. Nevertheless, it has not only an English existence independent of the combination d+y, but it cannot be shown to have been original to our tongue. The letterj spells it. The letter j, however, was strange to the A. S. alphabet; and equally strange were any combinations that may reasonably be supposed to have represented it. It can not only, then, not be shown to have been original to our tongue, but good reasons may be given for denying its presence in our earliest soundsystem. It is an evolved sound of comparatively recent origin, and, if we look at the words in which it occurs, we shall easily get the history of its evolution. It is rare in words of Angle, common in words of French and Latin origin; words which are sometimes spelt with a g (followed by e or i) and sometimes with a j. a. Gentle, general, gender, giant, gibbet, &c. b. Judge, just, juggle, jest, &c. Now, with all these words the initial sound in the original French was that of zh. Hence, whilst the ch in choose (tsh) has grown out of the A. S. k+y, it is out of the French zh that we have evolved our j, as in jest (dzhest), and our g, as in gibbet (dzhibbet). Not that the histories of the two evolutions, though different, are unconnected. The existence of the previous tsh of A. S. origin, undoubtedly, promoted the change from zh to dzh. Then there are the words in ch like chief, and chase, &c., &c., of French origin. The ch here originally represented the sound of sh. It now = tsh. The reason for this lies in the precedents and analogies of the A. S. k+y and its changes, and the French j (zh) with its conversion into dzh. I conclude my observations on this long and complex ques tion with remarking that, different as are the histories of our tsh and dzh (ch and j), they still illustrate the same general fact-viz. the instability of the combinations s, z, t, d, k, or g+y, and their tendency to become sibilant. With the case is With the ch and g (sh and zh) of the French, the very form of the combination suggests the history of the change. There was once a sound of k and g (as in gun). Otherwise the letters would never have been used as they are. less clear. The name, however, of "Jupiter" (writes Professor Key) "was undoubtedly written originally Diupiter, so Janus was at first Dianus, just as the goddess Diana was called by the rustics Jana."-Alphabet, p. 71. § 282. The chief points wherein the English sound-system differs from that of the more important modern languages, are worth noting; a knowledge of them being useful in the study of foreign tongues. The scarcity of proper open sounds contrasts the vowel part of the English sound-system with that of the Italian, French, and other languages. It is well known how common the sounds of both the a in father and the aw in bawl are there. In the French the e final is mute; so that the extent to which the open sound of the e in bed is wanting in English is not very manifest in the study of that language. Neither is in Italian, where no words end in -er. In German, however, and the Norse tongues, it requires some attention to discern the difference of sound between a final -e (as in meine), and a final -er (as in meiner). The absence of the é and ó fermé of the French and Italian, and other tongues, is another point to be remembered in the study of fresh languages. Thus the o in the Danish Kone runs great chance of being sounded by an Englishman as the oo in cool. The ❝ of the Germans (y Danish and u French) is a wholly new sound to the Englishman. So is the ö Danish and German, and the eu French. As these two sounds are both absent in Italian, Spanish, and Portuguese, the vowel-system of these languages is pro tanto more English than the French and German, &c. On the other VOL. II. C hand, the u in but gives foreigners trouble, being (as has been already stated) rare in the European tongues, though common in the Asiatic. In the simplicity of its nasal system (i. e. the sounds like the ng in king) the English agrees with the German, and is specially contrasted with the French and Portuguese. Wis English rather than continental. The best way for foreigners to learn it is to place an u- (oo in cool) before some syllable beginning with a vowel, and pronounce it as quickly as possible; e. g. on, u-on (00-on); et, u-et (oo-et), &c. In this the sound of w is soon obtained. way The mute-system in English is one of the fullest in the world. Out of the 4 quaternions 3 are full and perfect; so that 14 out of the 16 mutes belong to our language. The two that are wanting, the so-called aspirates of k and g, are the scarcest. Next to these come and þ, which we have. But though full, the English mute-system is simple. Each sound has its normal and typical form; so that the varieties which go by the names of guttural, cerebral, &c., are wanting. Hence the ch German and many similar sounds are strange to us. The question of simple single sounds is one thing; that of their combinations another. The diphthongs in English are simple and few; though some languages (e. g. the Swedish) have fewer. They are all, too, of one sort; i.e. that wherein the semi-vowel elements come last. Sounds like the French oi in roi we have none; nor yet (in the written language at least) any wherein y precedes its vowel. In the provincial dialects, however, they are by no means wanting. The nasal ng is never initial. We say song, but not ngos. This limitation of the nasal to the final parts of syllables is common. The Germans, Italians, &c., avoid an initial ng as much as does the Englishman. In the Keltic, however, it occurs, as it also does in many Asiatic languages. Though the English sibilants are compound, they are never complex. Thus, we say sha or sho. We also tsha or tsho. But we never combine the two; never use the complex sound say |