1849.] Sounds of the Language. 759 It is not yet known with certainty whether the Gallas have a written language, though the author of these volumes thinks it probable. The sounds in it are too numerous to be expressed with exactness by any known alphabet or syllabarium; but it was found that Roman characters could be best employed for the purpose, though in many cases only the sound could be represented by combining two or three letters. The number of characters, necessary to write the language, amounts to thirty. For a single consonant sound, which is rather a suspension of the breath than a sound, the Semitic Ayin () has been borrowed. The letters have, in general, the same sound which they have in the Roman languages. Thus a has, everywhere, the broad sound of a in fa ther; e has two sounds, one like a in fate; and the other like e in were, there; which sound is designated by the circumflex; as, méka, How much? I sounds like i in pin; o like o in no; u lik oo. All the vowels are, at the end of words, often almost mute, or spoken so short that they seem to be only a breath, aiming, as it were, at the sound which they denote; they are then written over the line, thus: Wac", Heaven; torban', Seven. They are all, likewise, either long or short. There are no real diphthongs, each vowel being sounded separately: k has always its hard sound, as in kill; c (written a before e and i) is a softer sound, produced by pressing the tongue against the palate, but without an aspiration; g is everywhere sounded hard, as in go; and never is melted, with n, into the nasal sound ng, so frequent in European languages, this sound being entirely wanting in Galla. Ch is used in only a few cases, as a euphonical softening of k before n and t in the inflection of verbs; as, nu bechna, We know; its pronunciation is the same as the German mich, dich. Tch has the harsh sound of tch in fetch; dj is pronounced as if it were dsh, but softer; and dy, sounded quickly, have their usual pronunciation. The T-sounds offer greater difficulties than all the other letters of the language the Gallas, however, make a very exact and sharp distinction between them, even when speaking rapidly. They are four : t is hard, as in tin; t is a very hard sound, peculiar to this language; it is formed by pressing the tongue closely against the upper foreteeth, so that after t a soft s becomes audible; dis our soft d in day, load; d' is very soft, and is formed by a gentle push of the tongue upon the hinder part of the palate; so that between it and the following vowel a slight pause seems to intervene, similar to the Semitic Ayin, thus resembling the sound of g, with which it is liable to be confounded. n is similar to the Italian gn in legno, or the Spanish ñ; z has the soft sound of s, and c the sharp sound; z' is the English sh. All the other letters have the same sound as in other languages, with the exception, perhaps, of b, which, in the middle of words, and particularly before vowels, sounds almost like v; thus d'abe is pronounced d'ave. The letters in the dictionary are arranged in families; method which, though more philosophical perhaps, is too foreign to our usual arrangement not to be a serious inconvenience in learning the language. In an Article like the present, a complete synopsis of the language will not be expected; we will, however, notice a few of its characteristics. "As almost in all languages," says the author (Grammar, page 9), "and especially in the Oriental, the verb is the soul of the whole and the root of nearly all the other parts of speech, so it is in the Galla language, although it belongs to none of the known families. For this reason, I have followed the custom adopted by Oriental grammarians, in placing the verb at the beginning. But in treating of it, I have, according to European usage, considered the inflection as the distinguishing characteristic of the conjugations; though, according to the views of Oriental grammarians, we might have inferred different conjugations from the different modifications a verb is capable of admitting, namely, by the use of affixes; (prefixes, in the ordinary sense, not occurring in Galla.)" The nature of the language fully justifies the plan adopted by the author. The Galla verb possesses in truth, so to speak, a remarkable productiveness. The national mind has, it would seem, been mainly directed to the multiplication of verbal forms, for the purpose of expressing wire-drawn shades of activity, to the neglect of many of the minor parts of speech. For besides its legitimate use, the verb is frequently made to perform the office of nouns, adjectives, adverbs, numerals, and negatives, whilst in particles and affixes denoting connection, relation, comparison, the language is so poor as to make us believe it is still in the earlier stages of development. All primitive verbs in Galla have the property of producing, by affixing syllables, new verbs, which are different modifications of the signification of the radical verb. The number of members belonging to each of the verbal families so produced is, however, extremely various, and depends on the nature of the radical verb; whence it arises that, in some verbs, singular forms are altogether wanting in the series; others are limited to only a few branches; and others, again, are capable of being extended to the sixth and even to the eighth link of the chain. We will illustrate this by an example : 1849.] Development of the Verb. bâ (root), To go out of a thing. 761 [pay. 11. bad'a, To go out for one's self, for one's own profit or damage. III. baza,To cause to go out, to let go out, to drive away, to pardon, to IV. bafad'a, To let go out, to drive away, etc. for one's self. v. baziza, To cause to let go out, drive out, pay, pardon. VI. bazifad'a,To cause to let go out, drive out,etc. for one's own profit. VII. baziziza, To cause to let drive out, pay, pardon. VIII. bazizifad'a. The same meaning, with the middle sense. In this it will be observed that the second, fourth, sixth, and eighth forms, ending in ad'a, are middle voices of the preceding forms, respectively; the third, fifth, and seventh have a causative signification. These causatives, however, depend for their precise meaning on the sense of the phrase in which they occur, and may be translated actively or passively; they give neuter verbs a transitive signification. The syllable za, which is the characteristic of the causative, is frequently extended; that is, the syllables ziza are appended, which, in distinction from the first, may be called a double causative syllable. The causative forms are regularly constructed, and proceed one from the other in a simple, uniform succession, according to the nature of the last radical. But the language does not stop here. As the branches of some trees take root and shoot up again like trees themselves; in like manner, the second forms become, as it were, new roots, endued with the same property of producing as the primitive roots, and thus new ramifications are developed, according to the same laws as in the beginning. In the same way, the iv. form (or second middle form derived from the first regular causative) gives rise to a new set of derivatives. Take, for example, mara, To turn round, to revolve; the simple forms of which are: mara. marad'a. marza. marfad'a. marziza. marzifad'a. marziziza. As to their signification, there exists but little difference between these causatives and those which are regularly constructed; the general manner of using these forms seems, however, to denote that the regular causatives imply a sense of "causing, operating; whereas, in the secondary (or middle-causatives), a permission only is expressed; so that, e. g. ñadz'iza signifies, To cause to eat, to feed; and ñad'ad'z'iza, To permit or allow to eat. The Galla has a great number of Onomatopoeticals, in which it is probably richer than any other language. The author distinguishes three classes: 1st, the language has the property of animating certain sounds or natural tones, commonly regarded as interjections, by affixing to them the terminations ad'a and fad'a, and thus confers upon them the nature, signification, and flexibility of verbs. Thus from hati, which imitates sneezing, it forms hatîfad'a, To sneeze. Such verbs resemble the middle forms, and are susceptible of the same changes. 2d, those verbs that already, of themselves, imitate such sounds, and consequently are to be taken as radical verbs; as, kakiza, to cackle, korriza, to snore. A 3d class comprehends all those formed by the combination of natural sounds with d'jeda, To say, and god'a, To make. In this manner any sound, or the tone of any noise, may be transfer. red into the department of verbs; and, consequently, there is a great number of them. It is interesting to observe in all these verbal formations, the endeavors of a people, still in the infancy of its intellectual development, to seize and render perceptible the most minute and fleeting things in nature. And not sounds alone, but motions and appearances are graphically represented. We select one example : dirgim djed'a is said of the sound produced when water meets over something which has been thrown into it: this natural sound dirgim, by taking the passive (reflective) ending ama, assumes the nature of a verb, having the signification of to wrap up, to sink down, to vanish; e. g. lafti dirgimame, The earth covered itself (with mist or darkness), darkness closed over the earth. Passive constructions are seldom used by the Gallas, though all transitive verbs may be easily changed to passive. They prefer speaking in the active; and if asked to translate passive sentences, they commonly avoid this construction as a difficulty, paraphrasing it with the 3d person plural active; e. g. instead of he was caught, they say, they caught him. The formation of the passive is very simple, being made, in radical verbs, by adding the ending ama to the root, and, in the transitive forms, by affixing the same to the last radical, which is, however, subject to certain alterations, which our space does not allow us to specify here. A peculiar feature of this language is the unequal development of its modes. It has, strictly speaking, but three: the Indicative, Imperative, and Infinitive. Tutschek includes the participle as a fourth; this is, indeed, very highly perfected, being provided with numbers and persons; yet it has rather the nature of a tense than a mode. No decided Subjunctive has yet been traced. It is true that, after certain conjunctions, in sentences where this mode would naturally be employed in other languages, the concluding vowel of some of the persous ends in u. But by reason of its deficiency in forms, and 1849.] The Galla Verb. 763 other imperfections, Mr. Tutschek does not call it a subjunctive mode, but prefers to designate it as the "mode in u.” In connection with the subject of modes, he says, p. 32: "There are, moreover, five mode-like relations, which are represented by affixed particles, and which, partly with and partly without the modification of the verb, may be appended to all syllables of flexion; and, either by themselves or by the aid of conjunctions, produce special shades of meaning. These particles are: re, mi, ti, yi, and in (ini); and as, up to the the present time, I have found nothing analogous to them in any other language, and being unwilling to invent new denominations, I call them, after the particles themselves, the mode in ti, the mode in mi, etc." These affixes present, in fact, some of the problems that remain to be solved by future investigators. Each tense, as also the present participle, is provided with a special form in the 3d pers. singular, to agree with feminine nouns. Collective nouns, to which the language has a decided inclination, are considered as feminine sing., although they comprehend masculine individuals. Hence the 3d pers. fem. of the verb is of frequent occurrence. Negation in the Galla is expressed sometimes by particles, and sometimes by negative verbs. Of the latter, there are three; they have this peculiarity, that, besides denying that an act was done, they imply the reason why it was not done. They denote, namely, that the subject either had not the will to do it, or he neglected it through inadvertence, or he had not the power. The indefinite future is formed by adding the auxiliary dyira, To be, with the infinitive; it is thus analogous to the English form, I am to love, or, I am about to love. Besides this, the Galla has still another method of expressing it more accurately, and in gradual approximation to the present; this is done by the tenses of the verb ga, To draw near, to approach; being combined with the Inf. of the verb in question, by which four different periods are distinguished, from the most distant future down to the time nearest to the present. "Thus besides ini d'ufufdyira, He will come, the Galla says also, as the time of the real arrival is more or less distant: 1. ini d'ufu gaufdyira, lit. He will approach to come; 2. ini d'ufu ga, He approaches to come; 3. ini d'ufu gae, he approached to come; 4. ini d'ufu gaera, or, gae dyira, He has approached to come; (when the arrival is close at hand;) and this precise distinction is strictly observed by the Gallas." Gram. p. 43. The Galla language has no article. The nouns are either primitive or derivative; the formation of those derived from verbs is, for the |