it is a corruption of conventio or coventio, which, in the oldest specimen of the Latin tongue, actually occurs in the sense of a general meeting of the people. If he wishes the authority of MSS., let him look to Sprengel's edition of Varro's De Lingua Latina. This editor takes great pains with his text, and always gives in the notes an account of the MSS. when they disagree from his text. Now he, like G. F., had a fancy for concio, which stands every where in his text, but as invariably the notes give "MSS. contio." Moreover, I support contio, if only to destroy the ludicrous etymology which would derive it from con and the verb cio, as though we had any right to avail ourselves of the first person of a verb in derivation, and as though the n of contionis, &c. could be neglected. 66 ... TUTTOVT, P. 112.-"The first person ending in om,"-" the analogy of am we cannot admit,"-" from premises so limited."—" The Greek verb," &c.—Strangely enough it happens that the Greek tongue affords the strongest support to my views. If we examine the middle or passive forms, τύπτομαι, τύπτεσαι, τύπτεται, . . . τύπο Tovtal, we soon discover that a is the distinctive characteristic of the voice, but this leaves us τυπτομ, τύπτες, τυπτετ, which agree wonderfully with the conjugation of a Latin verb, and speak decidedly in favour of a present first person in om, scribom. But G. F. perhaps believes the gulf between verbs in w and verbs in u to be impassable. All doubt seems to be removed by the optative túto μt, &c. Moreover, the German stock of languages furnish first persons, such as bim, "I be or I am," dom, "I do," gom or gam, "I go." Again, look at the plural τύπτομεν, or Doric τύπτομες. Of these the two last letters denote plurality, and we have left tuntop for the singular 1st person. Other support for this view is found in scribebam, scriberem, scripserim. The reason why the m disappears so readily after an o, it may be difficult fully to ascertain. But we have something very parallel to it in the absorption of the n in ratio, natio, and in the Portuguese habit of suppressing an n in the same position. Witness their orthography of their own capital, not Lisbon, but Lisboa. So the adjective fem. bona, becomes with them boa. I ought not to omit that the best Greek philologers believe ἔτυπτον, the 1st person, to be a corruption of ἔτυπτομ. See also my Grammar, 2 787. How readily a final m was lost in Latin is easily conceived, from the habit of treating one as nothing before a vowel in poetry, I mean in elision. Consi der too the practice of omitting one in writing, so that omniū means omnium. Consider further, how readily con becomes co, and that in the words consul, consolere, the n was treated as such a complete nonentity, that the inscriptions frequently omit it, cosul or cosol, cosolere, &c. The Greeks too, wrote the name of Constantine and the towns called after him Kwvstavivos, &c. with an o. Why? No doubt because the n was not pronounced. Hence the Cirta of Sallust in Numidia is now called Costantina, and Constantia and the Pagus Constantinus in France, are become Coutances and the Pays Cotantin. Observe too how readily in English none has been shortened into no. Such detailed proofs in a grammar would be wholly out of place, simply because of the room they require. P. 112.-" The nominative, the quarter FROM which," &c.G. F. has here forgotten, that as I deduce the passive through the middle or reflective, servus cæditur would signify originally "the slave strikes himself," so that the blow still proceeds from the slave. Thus, though I may be wrong, I am at least consistent. But the example in truth confirms my view, for the full phrase of the passive requires ab domino, whereas dominus would accompany the active voice. Thus dominus and ab domino express the same relation, viz. from, exactly what I contend for. P. 112.-" The nominative is formed by the suffix s.”—It is here objected that the exceptions are too numerous to be accounted for on one principle. Let us see. The exceptions may be reduced, I believe, to the three cases,-1. neuters; 2. nouns in a or the 1st declension; 3. nouns ending in a liquid, as r, l, n, or in ri or ro. To begin with the neuters, as my theory of the nominative insists on the case denoting the origin of motion, it must at first have been limited to living agents, that is, to masculine or feminine nouns, so that neuters had no original claim to the case ending. Secondly, although the Latin first declension has no s in the nominative, yet the Greek has, for example, noλity, 'Apiotaɣópaç, &c. But I may be told that this is limited to masc. nouns. This matters little, unless it be also contended, as by some it is, that the s is the symbol of the masculine. To such1 I again 1 The same parties are asked to account for this, that in acer, acris, acre, the form with an s is at least more feminine than masculine. object, that Homer freely uses nominatives without the for the masculine: Μητίετα Ζεύς, Ιππότα, &c. ; and that these very Greek nouns in a appear in the Latin prose writers of Cicero's age without an, viz., Aristagora, Archyta. But I go farther, and undertake to exhibit feminine nouns of this declension with an s in the nominative. It will be granted, I suppose, that copin and copía are but dialectic varieties of the same word. But this admitted, I claim the same admission for materia and materies, luxuria and luxuries, planitia and planities; and this the more, because I believe that Cicero declined the first of these nouns thus: N. materies, G. materiæ, D. materiæ, Acc. materiem, Ab. materia, thus blending completely together the so-called 1st and 5th declensions. The case of the liquids turns upon this fact, that the letter 8 is strongly influenced by the neighbourhood of liquids. The Greek verbs σπείρω, στέλλω, φαίνω, νέμω, have for their radical syllables σπερ, στελ, φαν, νεμ, and should therefore have formed the first aorists ἐσπερσα, ἐστελσα, ἐφανσα, ένεμσα; but by assimilation they appear to have become έσπερρα, έστελλα, έφαννα, ένεμμα. Next, the first of the consonants being omitted, the vowel was lengthened by way of compensation. Hence ἔσπειρα, ἔστειλα, eprva, evequa. Another example is in the form of Greek adjectives ending in liquids. The words TURTOVT, Xapevt, Kiλix, (I take them in their crude form shape,) made their feminines by the addition of oa, which by a slight corruption passed from tʊлτοντσα, χαριέντσα, Κιλικσα, το τύπτουσα, χαρίεσσα, Κίλισσα. But when this same feminine suffix was added to adjectives ending in p or, (I cannot call to mind any in λ or μ,) a modification took place. Thus, from μακαρ, μελαν, τερεν, were formed, first probably, μακαρσα, μελανσα, τερενσα, then μακαρρα, μελαννα, τε ρεννα, then μάχαιρα, μέλαινα, τέρεινα. 2 A third case in point.-Although the old Greek infinitive ended in ev or eva, the Romans in their oldest form appear to have had ese according to the usual analogy between the languages, (compare tónτou ε v, scribimus ov-oças-inus, &c.) Here then we have a suffix with an s, whence from es, "be," was formed an inf. esse, and from da, "give or put" an old inf. dase, (see dasi in Forcellini's Lexicon.) But this infinitival suffix ese becomes commonly ere, as in scribere, ferre; yet after the verb 2 Compare the middle form of äjjny and ägonv. vol or vel, "wish," the usual assimilation takes place, and we have velle. = A fourth case.-The Latin superlative should end in simo or sumo, but if the noun has an r or 7 in the last syllable, assimilation again modifies the s, and we have deterrumo, nigerrumo, facillumo, &c. In short a consonant r or l is very apt in Latin to dispose in a summary manner of a subsequent 8, and this no matter what the part of speech be. Thus for videbaris, we find videbare; for videreris, viderere, &c. Again, among the adverbs we might have expected tris, quatris, in agreement with our own twice, thrice, the Greek pis, nohλáms, &c., and the Latin bis duis. But in fact we find ter and quater. No wonder then if we also find, not linteris, puerus, paters, but linter, puer, pater. The Greek naτýp, to me seems to imply, by its long vowel, a previous form, πατερρ, from πατερς; and τέρην, a form τερεν» from τέρενς. But if the Greek nominative justifies such inferences, so also will the Latin pater,3 for that also is frequently long in the nominative. Precisely in the same way from the crude form pulvis, which already has an s independently of the nominative, (witness the diminutive pulvisculus contrasted with navicula, canicula, &c.,) was formed a nominative pulvis, for pulviss. See Virgil's line ending pulvis inscribitur hasta. Moreover, facilis, vigilis, strigilis, mugilis, have also the forms facul, vigil, strigil, mugil. No wonder, then, that consuls has become consul, originally perhaps consull. That rations should lose its 8, is in accordance with satin, viden, for satisne, videsne. N and S cannot co-exist in Latin; the one or the other gives way, at least in pronunciation. The tendency of liquids to assimilate is well seen in the old Norse tongue, where the comparative generally ends in ra, but yet vaen "fair," and sael "happy," make their comparatives vaenna, saella. Again, in the same language, the gen. plural commonly ends in na, yet dör "spear," makes its gen. pl. dörra. Compare also the Latin polliceor for pro-liceor, intellego for inter-lego, puella for puerula. P. 112.- Dative in bi.-Is the author at liberty to infer 3 Virg. Æn. v. 521, xi. 469, xII. 13; puer, Ecl. IX. 66; amōr, Ecl. x. 69; En. x. 872, XI. 323, XII. 668. 4 The old Norse (which usually forms its noun by the suffix r instead of s,) is not afraid of a double . Thus ketil, "a kettle," has nom. ketill, gen. ketils, from two or five words, &c." My arguments are founded upon a much wider base than tibi, sibi, ibi, ubi, alibi. First, I must include robis, nobis, utrobi; secondly, all the datives plural of the third, fourth, and fifth declensions in bus; thirdly, such datives of the 1st and 2d decl. as equabus, duabus, duobus, ambobus; fourthly, the Sanscrit datives dual and plural in bhy-am, bhy-as, the two last letters denoting plurality according to my view. Next, I find a great tendency in the b of Latin words to disappear, particularly after a short vowel. Tibi, sibi, ibi, ubi, are again and again monosyllables in Plautus and Terence. The same words have come down to the Italians, French, &c., as monosyllables. For instance, in French, te, se, y, ou, or toi, soi. Again, the imperatives abi, jube, are monosyllabic in Plautus and Terence. In the next place, the perfect and supine jussi and jussum, have no trace of a labial. The word amentum, too, seems by its meaning to be a corruption of habimentum, "something to hold by." At any rate, it must come from a verb, as its ending tells us. Besides, let this verb habeo be traced to the Italian and French, and the b will be found to have for the most part disappeared: J'ai, tu as, il a, ils ont. Even the 1st and 2d persons plural, lose their v, (the ghost of the b,) when they enter into the formation of the French futures, which it is now admitted are formed by the addition of the present of avoir to the infinitive of a verb: finir-ai, finir-as, finir-a, finir-ons, finir-ez, finir-ont. Nay, the conditional also is formed from avais, but loses the av. Thus, finir-ais, &c. Again, the Latin imperfect tense with suffix eba, soon lost its b when it descended to the modern tongues; recipiebam, receveva, recevea, recevoie, recevois, in which last word the v represents the p, not the b of the original word. (See Raynouard's Grammar of the Languages descended from Latin.) I have forgotten to mention the Homeric datives in, as Brypt, oùpavopt, ata@sapt. So much for the b, and its tendency to disappear. That the i belongs to every dative in its oldest form, I suppose I need not stop to prove. Putting all this together, I hold the base quite wide enough to support the argument; but the cognate "Tu te laves les mains." "Il se lave les mains." 6 So, also, from the German haben, to have, has a pres. hat, and a perf. 6 hatte. The Engl. also, has, had, without the b, or v、 7 So ab, sub, ob, lose their b in a, asportare; susum or sursum, sustollo, ostendo, omitto. |