UNIVLRSil V OF ILLINO'"" ' iBRARY URBANAChAMPAlGN STACKS The person charging this material is re- sponsible for its return to the library from which it was withdrawn on or before the Latest Date stamped below. Theft, mutilation, and underlining of books are reasons for disciplinary action and may result in dismissal from the University. To renew call Telephone Center, 333-8400 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN SEP 0 4 1986 L161— O-1096 Digitized by the Internet Archive in 2015 https://archive.org/details/publicschoollatiOOkenn_0 LATIN GRAMMAR. THE PUBLIC SCHOOL LATIN GRAMMAR FOR THE USE OF SCHOOLS, COLLEGES, AND PRIVATE STUDENTS BY BENJAMIN HALL KENNEDY, D.D. Nihil ex grammatica nocuerit nisi quod supervacuum est ; neque enim obstant hae disciplinse per illas euntibus, sed circa illas haerentibus. QuiNTiL. Inst. Orat. lib. i. cap. iv. NINTH IMPRESSION LONGMANS, GREEN, AND CO. 39 PATERNOSTER ROW, LONDON NEW YORK AND BOMBAY 1900 PREFACE TO THE SIXTH EDITION. § I. By an agreement between the Proprietors and Mr. John Peile, Fellow and Tutor of Christ's College, Cambridge, it is arranged that the ^Public School Latin Grammar,' with the books akin to it, shall hereafter be revised by Mr. Peile as joint and, in due time, sole Editor. Mr. Peile's merits as a classical and Sanskrit scholar, and as an able teacher, are widely known. Those who have the advantage of nearer acquaintance with him will be assured that the maintenance and improvement of these books, as means of public instruction, could not be placed in safer keeping. § 2. The First Edition of this Grammar, published in January 187 1, was introduced by the following Preface : — ' The " Public School Latin Grammar is simply a develop- ment of the Primer, in conformity with the design of those, who, after accepting the latter book, entrusted to the same Editor the preparation of the former. The difference between the elementary compendium and the higher work is such as might be justly expected. Whilst the general principles and many of the paradigms are in both the same, in the Grammar the subject-matter is arranged more systematically, the body of examples very much increased, the illustration wider, and a large amount of information is added, which in the Primer does not appear at all. DO vi Preface. ' Yet a Grammar of this size does not profess to be an exhaus- tive treatise on its subject. Competent and careful students, who combine its use with the reading of authors and the prac- tice of composition, so as to master its contents, ought indeed to become Latin scholars of considerable width and power; but they will still find much to learn in the field of Latin, which must be gathered from special monographies by eminent scholars, some of whom are occasionally cited in the following pages. ^At the present time, when the science of Comparative Philology has made such advance, that good living scholars know far more of the history and organism of the Latin lan- guage than was known to Quintilian and the old grammarians, the publication of a Higher Latin Grammar, without reference to the facts and principles of that science, would be a retro- gressive and senseless act. It must, however, be remembered that the chief end and aim of a Classical Latin Grammar is, to impress upon the minds of students the forms and construc- tions found in classical authors. Its office, therefore, is to use Comparative Philology as a guide and auxiliary in teaching Latin, not to teach Comparative Philology itself through the medium of Latin. This principle has been kept in view by the Editor throughout his work. The just mean is always hard to observe ; but he may venture to say that he has not strayed from it wilfully. In the Appendix, indeed, and in a few other places, he has thought it not inexpedient to cite some of the most important affinities between Latin and other Aryan languages and dialects ; but only with a view to point the path of future study, not to furnish the student with a sufficient knowledge of the several subjects there noticed.' 3. The following passages are taken from the Preface to the Second Edition, published in 1874 : — ^ Competent and candid critics are aware that a book of this size, in spite of its title, is not meant for school use in the same sense as the Primer and other lesson-books of a similar kind. As a school-book (for there is no limit to its use by any students who are capable of good private reading) masters can use it in two ways : (i) by enforcing general or occasional reference to its principles and rules in reading Latin authors ; Preface, Vll (2) by requiring definite portions to be prepared for periodical examinations conducted on paper as well as orally. ' The present opportunity has been used to enlarge and improve several departments of the Grammar, especially those of Soundlore and Derivation. To discuss the physiology of articulate sound has never entered into liiy plan. Were I competent to undertake this, which is not the case, I should hardly deem it suited to a book applying specially to Latin, but rather appropriate to a more general work treating of the Prolegomena to Grammar. * On the other hand, I have striven to bring out somewhat more prominently than before the leading facts of Comparative Philology, so far as they concern three kindred languages^ — Latin, Greek, and Sanskrit. The Sanskrit words iii this Edition are generally cited in their modern form. The term Primitive Root is, however, used; in what sense, and by what right, ap- pears in a Supplementary Note at the close of the Appendix. * As I am now, by the kindness of those whom I was bound to consult, authorised to attach my name to this Pre- face, I think it right to notice the chief objections made to the books on Latin Grammar with which I have been occupied. * When the Primer was published, seven years ago, it was right that it should be criticised, and certain that it would be impugned ; nor could we expect that all criticism would be equally candid and intelligent, or that every assailant would choose his weapons from the armories of truth and reason only. *The chief objection urged agaiiist the Primer was this : that it was too abstract and difficult for the use of children beginning Latin. There would have been some weight in this argument, if the purpose of the book had been rightly de- scribed. But it was really designed as a class-book, not for Elementary Schools and First Forms, but for all Forms in Public Schools below the grade of those boys who could pass with advantage to the use of a fuller Grammar. Other com- panion books were in preparation for the instruction of children at home or under private care j and these have since been published. ' It was, secondly, stated as a charge against the Primer, and subsequently against this Grammar (in which the teaching of the Primer is contained), that they " bristle with nev/, hard, and Preface. uninviting terms." This charge, urged as it has been with much persistence, and little concern for truth, must now be met by some remarks on the terminology of Grammar, together with a statement of my own feelings and practice in regard to it. § 4. ' Every science must have its own terminology. Grammar is a science; and in Latin Grammar, as one of its departments, there exist, I believe, more than three hundred technical terms. Most of these are either actually Greek words, as Syntax, Prosody, &c., or translated from Greek into Latin, as the names of the Cases and Parts of Speech. Others are purely Latin, as Gerund, Supine, Active, Passive Voice. Of these various terms, whatever the original unfitness of some, the larger number have struck their roots in literature so deeply and widely that any attempt to extirpate them w^ould be quixotic. Many, indeed, are in themselves unmeaning or inadequate (as Gerund, Supine, Deponent, Accusative, Genitive, Ablative); but the learner by gradual experience is enabled to use them practically, which is after all the end we wish to reach, though the road to it might at several points have been improved. A few terms, which are not only vicious, but really confusing, and at the same time unessential, I have exchanged for better substitutes. Among those so rejected are Neuter Verb, Neutropassiva, Neutral ia Passiva, Substantive Verb. Again, we find a considerable number of cumbersome Greek terms (Heteroclita, Heterogenea, Aptota, Diptota, Triptota, Tetraptota, with many of the names given to what are called Figures of Speech), which are of little use to learners. These may either be omitted, or, at least, dismissed to some unconspicuous corner. ^This statement affords ample proof that no* disposition existed to place in the student's hands a Grammar " bristling with hard and uninviting terms,'' though it is not unnatural to ask what those " inviting " terms are which, like the " crustula " of the " blandi doctores " in Horace's time, have magic power enough to attract young learners, elementa velint ut discere prima." ' But there is one important truth which many would-be critics either ignore or forget. Grammar is not only a science, but a science capable of constant improvement; and improve- ment in science usually brings with it some change in termino- logy, or some addition to it. Now, in every division of Grammar, — Soundlore, Wordlore, Syntax, and Prosody, — vast Preface, ix strides have been made in this century through the fruitful labours of scholars, chiefly German, some English ; whom I would gladly recount here, were I not afraid of omitting some name or names from so large a list. Accordingly it will be found by those who study the works to which I allude, that the terminology in each division has been more or less modified, more or less enriched. § 5. * As respects my own contributions to Latin Grammar, in the treatment of Soundlore and Wordlore I claim little origi- nality. If I have compiled judiciously and correctly from the works of great comparative philologers, so as to explain and illustrate usefully the received facts of Latin word-formation, I shall be amply satisfied with such credit. Again, in the Prosody of this Grammar I have no share beyond the Table of Metres and one of the Notes on Metre, containing little more than tabular enumeration. The rest I owe to the kindness of my friend Mr. Munro, whose recognised eminence as a scholar needs no praise from me to enhance it ' But the Analysis of Sentences (Simple and Compound) which constitutes the Syntax of this book, has been, to a great extent, the fruit of personal study, personal thought, personal labour. Sketched out in the Syntax of my " Elementary Latin Grammar," it is filled in, though far from reaching the fullness of perfection, in the present Grammar. * I speak from long personal experience when I say that any capable mind, which has fully mastered the principles of those pages (348-500, especially 348-359 and 434-500), will be able, in reading any part of Horace, Cicero, Livy, or Tacitus, to move through their longest periods with a firm intellectual step, realising, and, if need be, stating the 7'aison d'etre of every constructed word, especially (for this is the most crucial test) the raison d'etre of mood and tense in every Subjunctive Verb. The same mind, so prepared, and applying itself to write Latin, will be free from the risk of using any wrong construction. Not that the mastery of a grammatical Syntax alone will give the student stylistic power and skill in composition. These belong to the vis divinior^ to inspiration drawn by a gifted nature from the study of the best Latin authors themselves. To such study, combined with practice, no scholar will hesitate to assign by far the largest share in the formation of a good style whether of prose or of poetry. But, in the course of reading, the student X Preface. cannot afford to neglect any valuable help ; and of all appli- ances none is so valuable, none so indispensable, as a sound, well-arranged, and lucid Grammar. § 6. * The study of any language with its grammar contains more or less, according to the character of the language chosen, the study of every language and its grammar, the study of language in general and its grammar. The Greek and Latin languages (illustrated by their sister, the Sanskrit) are best adapted for this purpose, because their forms and constructions, themselves grand, are fixed in two grand literatures. One who studies these languages and their grammars cannot help studying to a great extent, coordinately with them, his or her own native language and its grammar. And the best mode and course of study will be that which is so conducted as to make such co- ordination as effectual and as widely instructive as possible. The principal reason why translation into Greek and Latin Verse as well as Prose deserves to be retamed in the practice of classical instruction I hold to be this, — that it is a valu- able exercise in the acquirement not only of those two dead languages, but of the learner's native living language at the same time. § 7. ' A book like the " Public School Latin Grammar " does not pretend to exhaust the subjects of which it treats — subjects on which many large volumes may be, and have been, written — but it carries the student very far on his road, and points and smooths the path of future acquirement. § 8. * I return to speak of my Latin Syntax, by which alone, so far as I know, my works on Grammar have obtained the favour and confidence of eminent scholars engaged in public instruction. ' The treatment of Latin Syntax has in the present century passed through a revolution scarcely less considerable than the treatment of Etymology. ^ The means by which this revolution has been wrought are: (i) the application to the whole doctrine of Syntax of the cor- relative logical terms Subject- Predicate and Subject- Object, with the principles they imply ; (2) the distinction between the Simple and Compound Sentence, and between the several kinds of each, with the consecution of tenses in them ; (3) the distinction between Oratio Recta and Oratio Obliqua, with the various affections which clauses subordinate to Oratio Obliqua receive. Preface, xi 'We owe to the perspicacity and learned labours of various writers, chiefly German, the reforms made in Latin Syntax. I cannot assign to each his due share. The Grote- fends, Kriiger, Zumpt, O. Schulz, Ramshorn, Kiihner, Madvig, Key, have each their special merits. Of these I place Raphael Kiihner in the first rank \ and I am much indebted to Grieben's "Lateinische Satzverbindungen." In cur own country the scholastic study of this part of Grammar was usefully promoted by the Exercise-books of T. Kerchever Arnold. ' These reforms brought into the teaching of Latin Syntax, besides the terms already named, a certain number more, per- haps from forty to fifty, including the names given to the seve- ral varieties of the Simple and Compound Sentence, with their subdivisions ; including also the terms Protasis and Apodosis in sentences which, like the Conditional, take these parts. § 9. ^ As regards the new terms which my own improvements have suggested, three alone have frequent and important prac- tical use ; the value of which I insist on as very great. These are, (i) Prolative (Infinitive) ; (2) Copulative Verbs, introduced first in my " Elementary Grammar " ; (3) Suboblique (clause or verb), a convenient abridgment of the necessary phrase Subordinate to Oratio Obliqua." ' Further, it appeared that the doctrine of copulative predi- cation in Grammar required, for its clear statement, the use of some terminology from which the term predicate itself should be excluded ; and this was at length found in the term used by Mr. C. P. Mason, (predicative) Complement. * I say then, generally, that a new term proposed in Grammar is not to be condemned because it is new; but, if at all, for one of three reasons : that it is superfluous ; or that it is in- adequate ; or because a better term is suggested. As respects myself, I repeat that I have not the least disposition to use hard terms ; and I say that those which I have introduced are unjustly so described. But I cannot adopt the poor pedantry which refuses to facihtate and abridge discourse by the use ot suitable terms ; to write, for instance (after due explanation) ''Collective Subject" rather than "Nominative Singular of a Substantive which implies a multitude of persons or things " : and "Composite Subject" rather than "two or more sLxigular Nominatives agreeing with one plural Verb/' XII Preface, § 10. 'My Elementary Latin Grammar," first published in 1843, obtained, after twenty years, approval so wide, that its circu- lation approached 8,000 copies annually : and, during those years, not a single complaint affecting its terminology was heard either from the public press or from the eminent teachers who used it in their schools. Such attacks broke out when it was adopted as the groundwork of a new school grammar; and their justice may be tested both by this fact, and by comparing the imaginary difficulty imputed to a few new terms in the Primer, with the many and great obstacles existing in its chief predecessor, Lilly's Grammar.' § 1 1. In the Preface to the Third and Fourth Editions certain portions of Syntax were discussed. Those discussions, being of signal importance to the right appreciation of Latin Com- pound Construction, will here be repeated generally: but with partial suppression of some topics and enlargement of others. I. The Doctrine of Predication. § 12. This Doctrine is treated (§§ 100-106) in agreement with the principles now received in all Continental Latin Grammars, and in most Grammars of the English language, but with some slight variations in the mode of treatment. Logic and Grammar are akin to one another ; but their spheres are different. Logic is the Grammar of reasoning : it develops ' the laws of thought/ Grammar is the Logic of language : it displays the rules and idioms of discourse. The Correlation and the Terms Subject- Predicate are necessary to both sciences. But the scope of these terms is not the same in both. If we take a Simple Sentence, such as * beneficium male collocatum nocet (noxium est) hominum societati,' we see that the Logical Subject of this proposition is 'beneficium male collocatum,' but the Grammatical Subject of the sentence is ' beneficium,' of which ' male collocatum ' is an adjunct. Again, the Logical Predicate is ' noxium,' the Grammatical Predicate ' nocet ' or ' noxium est,' of which ' hominum socie- tati ' is an adjunct. Hence appears the propriety and necessity (if confusion is a thing proper and necessary to be avoided) of distinguishing the terms Subject and Predicate in Grammar by the epithet ' Grammatical.' As for the terms Subject- Preface, xiii Predicate themselves, they have now so firm a footing in the science of Grammar that they cannot be excluded from it, if their exclusion were desirable. See * Predicate * in Index I. The Subject is 'id quod Praedicato subjectum est' : the Predicate is ' id quod de Subjecto praedicatum est/ The com- bination of the two (as Kiihner says : * Ausfiihrliche Grammatik der Lat. Spr.,' Part iii. § i) is rightly called the Predicative Relation, because the Predicate (or Verbal notion) is the kernel of speech, to which the Substantival notion stands in subjection, and is therefore called Subject ; often indeed expressed by the endings of the Verb (am-o, ama-s, &c.). When I was preparing my * Elementary Latin Grammar ' forty years ago, being in some dread of interference with Logic, I took for my type of simple predication, ' homo est mortalis.' But, when the Primer was compiled in 1866, the four Oxford scholars engaged in that work unanimously held that (in Grammar) Subject and Finite Verb are the true norm (homo morilur), and that Incomplete Predication (of the form homo est mortalis) should be taken afterwards as the large exception. This settled the question then, in accordance (as before noticed) with the practice of all continental writers : and a verdict thus authoritatively and generally pronounced is surely entitled to acceptance. II. Complement (of Predication). § 13. This suitable and useful term was first suggested by Mr. C. P. Mason in his ' English Grammar,' to designate that which completes the sense of a Simple Sentence when the verb is one *of incomplete Predication' (called 'Copulative' in this Grammar, p. 351). In sentences such as 'homo moritur (est mortaHs),' we have seen above that the Grammatical Predicate is (not 'mortalis,' but) 'moritur' or 'est mortalis.' Donaldson's expedient, of using the terms ' primary, secondary, tertiary ' predicate, I cannot approve. It confounds confusion, invades the domain of Logic gratuitously, and carries into the rules of Grammar the use of a word (predicate), which, however necessary to the preamble of Syntax, as the correlative of Subject, may be replaced afterwards by the term Finite Verb (or Verb of the Sentence) with great advantage. All confusion is happily avoided by the term ' Complement,' which is wide xiv Preface. enough to include every word or phrase capable of completing the construction of a Copulative Verb, whether finite or infinitive. See the Examples on p. 352. III. Relations in the Simple Sentence. § 14. Mr. Mason, in his ' Engfish Grammar,' following Becker's ^ Organism der Sprache,' treats of the Relations of Words in the Simple Sentence. The ^ Public School Latin Grammar ' does the same. One of our critics regards these Relations as * spurious children of Logic and Grammar.' But he has failed to interpret the procedure rightly. It is as purely grammatical as any procedure can be, which admits (what no grammarian can now exclude) the correlations Subject-Predicate and Subject-Object. Two of Mr. Mason's ' Relations,' the Predicative (I.) and the Objective (HI.)? ^^e the same, in title and extent, as those of this Grammar. His ' Attributive ' Relation contains the Qualitative (IL), but is more extensive: his 'Adverbial' Relation contains the Circumstantive (V.), but is more extensive. Mr. Mason was dealing with English, a language of rare inflexions, using Prepositions in their stead. I deal with Latin, a largely inflected language. But even in English the Genitive should not be merged in the Attributive Relation, and the Dative Case in the Adverbial (Circumstantive): much less in Latin. For, true as it is that numerous instances of the Genitive are attributive in character, and that many Datives might be replaced by Preposition with case (i.e. adverbially) ; still there remain very many examples of each case which cannot be so represented, and this fact, combined with that signal distinction between forms of construction, which merits distinct treatment in Grammar, leads to the conclusion that the Dative and Genitive Cases ought to rank as separate Relations. The Dative is therefore classed here under the * Receptive' (IV.), and the Genitive under the ' Proprietive ' Relation (VI.). Relation VI L, that of ' the Prolative Infinitive,' appeared for the first time in the * Pubhc School Latin Primer.' It com- prises all the instances in which the Infinitive extends (profert) the construction of words capable of being followed in de- pendence by a Copulative InfiniUve with Nominative Com- Preface, XV plement. See § i8o. In the 'Elementary Latin Grammar' the Infinitive with some of these Verbs (soleo, possum, &c.) was called Objective ; with others (videor, dicor, &c.) Predi- cative (/>. complemental). But these shifts never satisfied: for if, in 'soleo errare/ the Infinitive is Object of ' soleo/ it is an unique Object : and if, in ' videor errare,' the Infinitive is predi- catively complemental (which in some sense it is), its character as a ' Complement ' is widely distinct from that of an Adjective or Substantive (which qualify the Subject), and from every other instance in p. 352. And how, on the same principle, can we analyse without the most unpleasant confusion such sentences as these? — • Marcus putatur velle fieri philosophus. Sapientis est velle fieri doctiorem. At length a conviction was reached, that this usage of Grammar (common to all Aryan languages at least) deserves separate classification as a specialty of the Infinitive Verb-noun. Madvig's mode of treating this construction is not essentially different in principle. Under one head (§ 180) this Grammar gives what he sets forth in three places (§ 389, § 393, § 400). He treats in one and the same chapter of the Infinitive in Oratio Recta and Oratio Obliqua. Deeming it right and important to keep Simple and Compound Construction apart, we consider in Chapter III. the Infinitive of Oratio Recta, in Chapter IV. that of Oratio Obliqua. But when Madvig speaks of the Infinitive as joined to these (extensible) Verbs *um den Begriff zu erganzen und die Handlung zu ergeben ' (to complete the idea and supply the action)^ this is exactly what is meant by the Prolative Relation of the Infinitive : and it is very much the same as the use of the Infinitive, in German and English, with those Verbs which some grammarians have very inade- quately called 'auxiliary' (ich will, soil, kann, muss, &c. kommen : I will — shall — can — must, &c. come). The con- struction belongs also to French, a Romance (latinistic) language. For though French inflects (with Latin) / will come^ I would come, by 'je viendrai, je viendrais,' it falls in with Latin, German, English, in saying je peux — ^je veux — ^je desire — j'ose, &c. venir. It is unquestionably true that after many of these Verbs the Infinitive may be called an Object by anybody who wishes to do so, as in 'vincere scis, tu sais vaincre/ ' cupis abire, tu desires partir,' &c. The use of the xvl Preface. Verb noun as an Object is recognised in § 179. But the reasons in favour of accepting a distinct Relation wherever the test of * esse/ &c. with Nominative will apply are decidedly preponderant. If an example be adduced like this in Horace, C, i. 2. 49 : — hie magnos potius triimiphos hie a m e s did pater atque princeps^ and if it be asked whether, as triumphos\% Object of ames, did does not also stand in the same construction, the reply is — that Latin v/riters, especially poets, often construct one Verb with dependence of two kinds : so Verg. Aen. iii. 234 : — sociis tunc arma capessanf edico et dira bellum cum genlt gerendum, * Ames,' in the lines of Horace, first takes an Accus. Object frtumphos, and then a Prolative Infin. did^ with its comple- nients. The example belongs to that kind of construction which grammarians have called Zeugma. See § 61. Our last Relation, the Annexive (VIII.), is in kind dif- ferent from the other seven. It is really no more than a com- pendious method^ by which a word B is noted as assignable to the same Relation with a preceding word A. It is a con- venient substitute for those cumbersome and yet incomplete rules which in the old School Grammars were meant to account for the cases, moods, &c. of words linked to others by various conjunctions. See Supplementary Note II. p. 579. § 15. There are two great facts in Grammar which the student of language should always bear in mind : — (1) Few Definitions are free from examples which occa- sionally stray beyond the precincts there laid down, to enter those of another Definition. For instance : a Substantive may sometimes become an Adjective (rex, regina, raptor, victor, victrix, &c.) : an Adjective or Participle often becomes a Substantive (sapiens, utile, utilia, adulescens, sponsus, dictum, &c.) : a Verb contains a Noun among its forms : a Noun sometimes takes the functions of a Verb : an Adverb becomes a Preposition, a Preposition an Adverb : Declensions encroach V?pon one another ; and so on. (2) A Norm or Rule may be liable to numerous exceptions: Preface, xvii and yet, even if the excepted instances could be shewn to equal or even exceed in number the instances which obey the rule, the Norm ought to remain paramount, and not to be extended in order to recognise such instances as normal. See § loi. Thus, referring to (i), all Annexed Words belong to some one or more of the other Relations also. Every Complement, belonging, as such, to I. will fall under some other Relation also. Of those which occur in the examples, p. 352, the first six fall under II., the seventh and eighth under VI., the tenth under IV., the ninth and eleventh under V. Most examples of Relation VI. and some of IV. V. are akin to II., being attributive in sense, but excluded from II. because thev appear as caseforms, and not in attributive concord. § 16. The foregoing observations shew that, in the mode of treating these Relations, there is no spurious intrusion of Logic into Grammar. The Dative is not merged in the category of Circumstance, nor the Genitive in that of Attri. bution (Qualitative). Each case has its own sphere : the Nomi-. native (as Subject-case) and the First Concord are in I., the Accusative as Case of nearer Object is in III., the Dative as remoter Object-case in IV., the Ablative (with the Accusative depending on Prepositions) in V., the Genitive in VI. The Concords 2, 3, 4, come under II.; the peculiar use of the Infinitive under VII., the linking by Conjunctions under VIII. Afterwards, the Vocative and all Interjectional usages lying out of the Sentence are separately treated, and then the theory of the Relative. Grammar is followed, Grammar kept in view, throughout. Experience proves that such a synopsis of the Simple Sentence does materially help many students to read with more profitable appreciation the rules that follow, and, reviewed again at the close, will map the subject in their niind3 more lucidly and more enduringly. IV. Ellipse of the Finite Forms of *Sum.' § 17. This topic is considered in the note on p. 428 : see also § 99, Munro on Lucr. ii. i, with the authorities there cited. The ellipse occasionally creates misinterpretation, participles finitely used being sometimes mistaken for mere participles (Hor. C. i. 37. 25; ii. 9. 15), and again mere participles having been regarded as finite : thus in Verg. B. ii. 40 : — a xviii Preface, praeterea duo nec tuta mihi valle reperti capreoli, sparsis etiam nunc pellibus albo, bina die siccant ovis ubera; quos tibi servo. Wagner and Ribbeck have a semicolon after *albo/ thus apparently making ' reperti ' finite : but the tenour of the passage indicates that ^capreoli siccant' is the principal predi- cation, and ' reperti ' a mere participle. The ellipse of * esse ' in oblique construction, when the par- ticiples perfect, future, or gerundive occur as accusatives in dependence, is familiar to every reader of Latin authors in prose and poetry. But the Prolative construction, by which the Nominatives of these Participles are used as Infinitives without * esse,' is not by any means so generally and so well understood by young scholars. They are therefore advised to study with care the note on § i8o in this Grammar, and to collect other examples of this construction (the Participles in the Nominative as Infinitives without ' esse which are not duly recognised in Madvig's Grammar. It may perhaps be more clearly exhibited by setting side by side the Accusative (Oblique) and the Nominative (Prolative) constructions in a few examples. a. T. Manlium locutum ferunt, T. Manlius locutus fertur, Liv. b. Ferunt Promethea coactum . . . Fertur Prometheus coactus . . . Hon c. Delectum habendum putant, Delectus habendus putatur. d. Omnes secuturos verisimile est, Omnes secuturi videntur. e. Graeciam collisam narrant, Graecia collisa narratur, Hor. f. Memorant quendam solitum . . . Memoratur quidam soHtus, . . . Hor. g. Credimus Athon velificatum, Creditur Athos velificatus, luv. h. Perunt genetricem adfatam lovem, Fertur genetrix adfata lovem, Verg. ix. 82. In every one of these examples ^ esse ' is to be mentally supplied — its construction being Oblique (§ 194) in each former — Prolative (§ 180) in each latter line. Preface, xlx V. § 1 8. Some nice points of Syntax have been either over- looked or inadequately treated. Such are the Substantival constructions with ut and quod, in place of an Infinitive Clause. See §§ 195-6. Still more unfortunate has been the treatment of constructions ranked in this Grammar under the title Petitio Obliqua, § 197. A disposition is shewn by some writers to make these mere varieties of the Adverbial (Final) Clause with ut, ne, although their prominence and importance in Narratio Obliqua (§ 230) prove their just rank as one of the three varieties of dependent Substantival Clauses, which con- stitute Oratio Obliqua. The Statement (Accusative and In- finitive) and the Question assert themselves, as it were, and cannot be ignored : but the Dependent Petition has to wage a sort of rivalry with other constructions of u t, n e, in order to obtain its just place in Grammar, as representing an Oblique * permission, command, or request.' The examples which Madvig cites in §§ 403-4, shewing the juxtaposition in Narratio Obliqua of indirect statements, commands, and questions, might have shewn him the true order in which the three ought to be treated. ^ Si pacem populus Romanus cum Helvetiis faceret, in earn partem itu7vs atque ibi futuros Helveiios, ubi eos Caesar cpn^ stituisset atque esse voluisset : sin bello persequi perseveraret, reminisceretur et veteris incommodi populi Romani et pristinae virtutis Plelvetiorum . . . quare ne committeret ut is locus, ubi constitissent, ex calamitate populi Romani . . . nomen caperet.' — Caes. B, G. i. 13. ' Cum vellet, congrederetur ; tntelkcturiiin quid invicti Germani .... virtute possent' — Caes. B. G. i. 36. * Duces pronuntiare iusserunt : " ne quis ab loco discederct \ illorum esse praedam atque illis reservari quaecumque Romani reliquissent : proinde omnia in victoria posita existimannt,^^' — Caes. B. G, v. 34. ' Cicero respondit : " non esse consuetudinem populi Romani accipere ab hoste armato condicionem : si ab armis discedere velint, se adiutore utantur legatosque ad Caesarem mittant ; sperare^ pro eius iustitia quae petierint impetraturos." ' — Caes. B, G. V. 41. ^ Nuntia Romanis : " caelestes ita velle ut mea Roma caput or bis terrarum sit : proinde rem militarem colaiit ; sciantque 2l z XX Preface. et ita posteris tradant^ nullas opes humanas armis Romanis resistere posse."' — Liv. i. i6. Exprobrant multitudini : " saginare plebem populares suos, ut iugulentur. hoccine patiendum fiiisse, si ad nutum dictatoris non respondent vir consularis ? fingerent mentitum ante, atque ideo non habuisse quod turn responderet : cui servo umquam mendacii poenam vincula fidsse ? " ' — Liv. vi. 1 7. * Blaesus multa dicendi arte, " non per seditionem et turbas desideria militum ad Caesarem fe7'e?ida,'^ ait ; " neque veteres ab imperatoribus priscis neque ipsos ab divo Augusto tarn nova petivisse \ et parum in tempore incipientes principis ciii-as onerari : si tamen tenderent in pace temptare quae ne civilium quidem bellorum victores expostulaverint, ciir contra morem obsequii, contra fas disciplinae vim 7neditenhir ? decemerent legatos seque coram mandata dareiitT ' — Tac. An7i. i. 19. * Eo in metu arguere Germanicum omnt^, quod 7ion ad supe- riorem exercitum pergeret^ ubi obsequia et contra rebellis auxi- lium : " Satis superque missione et pecunia et mollibus consultis peccatum ; vel si vilis ipsi salus, cur filium parvulum, cur gravi- dam coniugem inter furentes et omnis humani iuris violatores haberef^ illos saltem avo et reipublicae reddereV^^ — Tac. Ann. i. 40. See do. do. ii. 15. * post paulo scribit sibi milia quinque esse domi chlamydum ; partem vel toller et omnes.' Hor. Epist, i. 6. 43. Compare Verg. Ae7i, iv. 683 : date volnera lymphis Abluam^ &c. A en, vi. 884 : manibus date lilia plenis Purpureos spargaTn flores, &c. The true construction, Mate abluam,' g7^a7tt 7?te to wash away^ ' date spargam,' grant me to scatter^ &c., has in each place been recognised fully by no commentator except Ladewig. If commentators v^ho have fallen into error respecting them had been familiar with the principles of * Petitio Obhqua,' they would have seen that the Subjunctives depend as Objects on ' date,' like ' colamus ' in the following lines of an Inscription to Silvanus found at Aime in France : tu me meosque reduces Romam sistito daofi.^ Itala rura te colamus praeside. — Coll. Orell. 1613. Preface, xxi Had ' sinite ^ been written instead of ^ date ' (and there is no real difference) the mistake would not have been made. VI. § 19. Chapter VI. of Part I., Division ii. (§§ 61-99), on the Uses of Words, though subjoined to Wordlore, may be read by those who have already studied Latin Syntax to some extent in a shorter Grammar with suitable practice. It unavoidably contains many topics (as, Ellipsis and Zeugma, Prepositions, Correlation, Mood), which belong in principle to the construc- tion of Sentences, and which many grammarians, as Madvig, intermingle with the rules of Syntax, thereby, we think, sadly breaking the continuity, and obscuring the doctrine of these latter, as intended to develop the construction of Sentences, Simple and Compound. To those who study this Grammar we strongly recommend the adoption of the following order, in studying the laws of Words constructed in Sentences; i.e. Syntax. (1) Wordlore, Division ii., Chapter VI. , Sections i.-viii. (§§61-89). (2) Wordlore, Division ii., Chapter II., Section x. (Numerals, §§ 32-34). (3) Syntax, Chapters I., II., III., IV., Section i. (§§ 100-189). (4) Uses of the Verb (Wordlore, §§ 90-99). (5) Syntax continued (§§ 190-250). The whole Chapter on the Uses of Words may be reperused with advantage at the close of such a course. VII. § 20. The systematic order in which the Doctrine of Sentences is drawn out is the chief characteristic feature of this Grammar. Chapter I. of Part II. (§ 100) sets forth : (1) The distinction of Sentences as Simple or Compound. (2) The three forms of the Simple Sentence : Statement (enuntiatio) : Will-speech (petitio) : Question (interrogatio). (3) The forms which these three severally take when, being subordinated in compound construction, they become Substan- tival Clauses : — Oblique Statement : Oblique Will-speech : ObUque Question. Chapter II. (§§ 101-106) contains : The Analysis of the Simple Sentence, and the eight Rela- xxii Preface, tions comprised in it : adding to these the Interjectional use of the Vocative and other Cases similarly interposed. Chapter III. (§§ 107-188) contains : Rules and Examples of construction in the Simple Sentence (Agreement : Cases : Infinitive with Gerunds and Supines). Chapter IV. (§§ 189-240) treats of : The Compound Sentence, in five Sections. Section I. takes up the topic begun in Chapter I., and shews: (1) Subordinate Clauses, of three kinds ; Substantival (§ 100) : Adverbial : Adjectival : (2) Adverbial Clauses, of seven kinds : (3) Adjectival Clauses, being in some kinds substitutes for Adverbial (see § 204). Section II. states the laws of Mood in subordination to Oratio Obliqua, actual and virtual, with examples. Section III. contains : Rules and Examples of the construction of the three varieties of Substantival Clauses : (1) Oblique or Indirect Statement (enuntiatio obliqua). (2) Oblique or Indirect Will-speech (petitio obHqua). (3) Oblique or Indirect Question (interrogatio obliqua). Section IV. contains : Rules and Examples of Adverbial and Adjectival Clauses stated in connexion with one another. Section V. forms a Supplement, treating of : (1) Consecution of Tenses : (2) Narratio Obliqua : (3) Reflexive Pronouns in Clauses : (4) Participial Construction. VIII. § 21. The scheme of Latin construction thus shewn forms an edifice of its doctrine, from which no stone can be taken away or displaced without damage to the whole fabric* * Let me here state my meaning more distinctly. I. I consider it desirable that the Uses of Words and the Rules of Construction should be kept generally distinct : but I deem it essential that the Uses of the Verb and the Doctrine of Moods should be learnt before the Laws of Compound Construction. This opinion is illustrated in Appendix IL to the ' Public School Latin Primer ' (years 1878 &c.). Preface. xxiii Nor can a single fact or principle laid down in it be denied by any one who is able to recognise facts in language, and to deduce principles from them correctly. (1) The Simple Sentence has three Varieties : I. Vales: 2. Vale: 3. Valesne? Can this be denied ? (2) Each Variety can be subjoined (with some formal change) to a principal Predication ; such junction being a ' Compound Sentence/ the subordinate or dependent member in which we term ^ the Clause.' 1. Audio (constat) te valere, 2. Opto (optandum est) ut valeas. 3. Quaero (quaeritur) (die) valeasne. Can this be denied ? (3) Each of these Clauses is related as Object or Subject to the Verb on which it depends. 1. I hear (it is evident) — What? That you ai'C well, 2. I wish (my wish must be) — What ? That you be well, 3. I ask (it is asked) (say) — What? Whether you are well. Can this be denied? (See it even in Madvig, § 208b, 398a.) 2. It is essential that Syntax should take for its starting-point the three forms of a simple sentence and their transformation into clauses when they become subordinate. 3. It is essential that the study of Simple and that of Compound Sentences should be treated in distinct parts of the Grammar, and that the rules for the Simple Infinitive, with Gerunds and Supines, should be in- cluded under the Simple Sentence, leaving the Infinitive Clause (though it may be cursorily mentioned) to take its proper place at the head of Compound Construction. 4. It is essential that the Doctrine of Compound Construction should take for its starting-point the threefold distinction of Substantival, Adverbial, and Adjectival Clauses, shewing the intimate relation of the two latter classes 5. It is essential that the Substantival Ut-clause and the Substan- tival Quod-clause should be shewn in their true place as succursal to the Infinitive Clause, with due notice of the relation which they severally imply to Consecutive and Causal Construction. 6. It is highly important that (while the term Conjunctive is given to the Mood generally) the term Subjunctive should be confined to its subordinate use 7. The distinction of Compound and Complex Sentences, which some English grammarians use, is superfluous in Latin. That of Coordinate and Subordinate Clauses (§ 100) answers the purpose adecjuately. XXIV Preface. (4) Therefore each of these Clauses has the property of a Substantive, and is justly termed * Substantival.' Can this be denied? (5) Of other Subordinate Clauses, those which answer the adverbial questions — ho7v^ for what purpose, why^ when, on what condition, &c. — are justly termed Adverbial Clauses (Con- secutive, Final, &c.). See § 189 B. There can be no just reason to deny this. [All Relative Clauses — formed by qui or a particle expli- cable by a case of qui, as quo, unde, cur, &c., may be called Adjectival, having the attributive nature of Adjectives. But in this Grammar (§ 189 C, § 204, &c.) the term is applied only to those Relative Clauses which determine Mood to be Subjunc- tive : such as : Quis est t a m i m p i u s qui parentem feriat ? = ut feriat ? — Missi sunt qui specularentur = ut specularentur : — Miseret me tui qui tantum desipias = quum . . . desipias : and the like. The larger power of the term Adjectival, as belonging to any Relative Clause, should not be forgotten, though its use is needed here alone to complete the analogy.] This classification of Clauses, as Substantival, Adverbial, and Adjectival, is recognised by the soundest German gram- marians, Kriiger, Kiihner, Feldbausch, Grieben, and many others. Its omission is among the chief faults of Madvig's Syntax. (6) Returning to Substantival Clauses (2), we observe that €ach of the Clauses is indirect, i.e. dependent on a Verb, which is itself dii'ect, i.e. independent. This indii'ect speech is called by general consent of grammarians ^ Oratio Obliqua,' and that on which it depends is called ' Oratio Recta' (direct speech). Thus it appears, that all three forms of the Substantival Clause constitute Oratio Obliqua. This is allowed, though haltingly and inadequately, by Madvig, § 403, Obs. i. The chief reason why oblique statement (te valere) has been ' specially called' Oratio Obliqua is this : that ordinary discourse in prose consists mainly of statements. Another reason is, that the indirect expression of the Imperative (bidding-speech or will- speech) is not so manifestly distinguishable from other forms as the Infinitive Clause (te valere), about which there can be no mistake. See what is said above of Petitio Obliqua. The student is strongly advised to keep this larger sense of the term Oratio Obliqua always in mind, and to fortify it by careful Preface, XXV study of Oblique Narration, as used by Livy, Sallust, and Tacitus. He may also consult with advantage the Syntaxis Vergiliana in our edition of Virgil, pp. 664, &c. (7) The Hmits of Oratio Obliqua being thus established as coincident with Substantival Clauses, we pass to the Mood of Verbs in subsequent Clauses depending on them, which we therefore call 'Suboblique,' that is, * Subordinate to Oratio Obliqua.' The rules on this subject are given in §§ 190-193, because the constructions resulting from them occur in many of the examples cited in the sections following. The Conjunctive is, by its nature, the Thought-mood or mood of conception. Hence, when a finite verb in secondary depen- dence forms part of the same conception as the Oratio Obliqua in primary dependence, it is put in the Subjunctive (dependent Conjunctive). See Example in § 190 L So also : Apud Hypanim fluvium Aristoteles aitbestiolas quasdam nasci, quae unum diem vivant^ Cic. T. D, i. 39. Perspicuum est, non esse utilia, quae sint turpia, Cic. Off, iii. 32. With the other examples on p. 437 of this Grammar, and those in ^Public School Latin Primer,' p. 167. This doctrine is laid down in all Latin grammars. IX. § 22. So also Rule 193, which states that a subjunc- tive is used in dependence on another Verb in the Conjunc- tive Mood, is in the nature of a corollary to Rule 190, and is not disputed. Rule 191 relates to implied or virtual Oratio Obliqua. The doctrine on this subject I have somewhere seen described as a mystery, too abstruse for anybody to understand or study. Now the differential calculus, or loga- rithms, or even decimal fractions, remain a mystery to those who have not taken the trouble to learn them. But Grammars are written for those who are willing to learn, and who wish to know well what they profess to know at all. It seems, therefore, that a few words here may not be wasted in the endeavour to clear up a subject which, after all, has nothing in it mysterious. For this purpose, it is best to begin with the simplest obvious examples. Compare, then, the two fol- lowing places in Cicero's Treatise * De Officiis ' : (1) Cyrenaici . . . virtutem censuerunt esse lau- dandam, quod efficiens esset voluplatis, iii. 33. (2) Laudat Africanum Panaetius, quod fuerit abstinens, ii. 22. xxvi Preface. We say that ^ virtutem esse laudandam ' (actually) is Oratio Obliqua, on which ^quod esset' depends, and is therefore Subjunctive, being ^ suboblique.' We say that ' laudat ' (virtually) contains Oratio Obliqua, and that * quod fuerit/ depending on it, is ^ virtually suboblique.' Such is our proposition. Let us consider it. First, as to * quod esset' in sentence (i). ^ Virtutem esse laudandam ' is Oratio Obliqua in its principal form of Accus. with Infin. (Infinitive Clause), and a Finite Verb really depending on such a form will be Subjunctive because the reason given for virtue being praiseworthy as well as the fact itself is referred to the mind of the Cyrenaics, and for this purpose the Thought-mood (Conjunctive) is employed. Such is the rationale of a Subjunctive * actually subordinate to Oratio Obliqua,' or (for brevity's sake) * suboblique.' Secondly, as to * quod fuerit,' in Sentence (2). Do we utter * a mystery ' when we say, that a person who is said to praise another, is said to think and to express some- thing ; that * laudat * necessarily contains the meaning * putat esse laudandum ' with the meaning * ait esse laudandum ' ? Enough that it contains the latter. Laudat then contains * virtual (i.e. implied) Oratio Obliqua ' : and the Finite Verb depending on it (quod fuerit), being really subordinate to a virtual Oratio Obliqua, or (for brevity's sake) * virtually suboblique,' is referred to the mind of Panaetius by becoming Subjunctive. He gives the reason why he praises. Such is one of the simplest instances of ' virtual Oratio Obliqu^-.' X. § 23. Here it will be right to deal with a plausible objec- tion, which may lead some not unintelligent minds to question the merit of the terminology used. Why, they may perhaps say, is a term which itself needs explanation, and which suggests a merely formal cause, interposed between the learner's understanding and the true logical reason of the Thought- mood, viz. that it refers the proposition to the mind of the Subject? The answer to this objection has already been suggested in another part of this Preface. Every science is taught and learnt through the medium of terms. It is the teacher's busi- ness to see that his pupils do learn — do know — the meaning Preface, xxvii and force of such terms. It is a learner's business to acquire their meaning and force, either from his teacher (if he has one) or from his books (if he studies privately). If he uses terms, of which he has not learnt the true meaning, he walks in the dark, and the results can only be ignorance and error. A good teacher will be always on his guard against this danger. If he asks a question, and is answered by a correct term, which he is sure the learner understands, he may say ' quite right,' and pass on. If he doubts this, he should cross-examine. For instance, As to passage (i) : Q. Why is ' esset ' Subjunctive ? A, It is suboblique. Q, How so ? A, It is subordinate to the Oratio Obliqua * virtutem esse laudandam.' And this Oratio Obliqua itself? A, It is subordinate to the principal sentence 'Cyrenaici censuerunt.' To what then is the Clause ^ quod efficiens esset volup- tatis' referred? A, To the mind of the Subject Cyrenaici. As to passage (2) : Q, Why is ^fuerit' Subjunctive? A, It is virtually suboblique. Q. How so? A, It is subordinate to an Oratio Obliqua implied in * laudat.* Q, How would you express this Oratio Obliqua? Ait esse laudandum (or some equivalent). Q, To what then is the Clause ^ quod fuerit abstinens ' referred ? A, To the mind of the Subject Panaetius. If the question were in class, and the catechumen failed to answer, the teacher would probably explain publicly, and re- examine privately, till he was sure the matter was understood. If our imaginary disputant, returning to the charge, says : May not this cross-examination be cut short? is not all con- tained in the last question and answer ? No, we reply : for we are not teaching Logic only, but also Latin : Latin construction, xxvlii Preface. Latin procedure, with its rationale. The attempt to teach the rationale without the forms which lead to it would be a double failure : grammar would manifestly be sacrificed, and Logic (we believe) would gain nothing by the sacrifice. XL § 24. Some persons imagine they have solved all Hhe mystery' of such constructions as (2) by saying that * The Sub- junctive is used in Causal and Relative Sentences to denote an alleged reason or act.' These words we quote from one such writer. ^ ^ Causal and Relative Sentences ' certainly do (for obvious reasons) supply the most numerous instances of ' virtually ^ It may be instructive to cite this writer's * ipsissima verba,' as an in- stance of error growing out of the attempt to defend error. He says : *The Subjunctive is also used in Causal and Relative Sentences to de- note an alleged reason or act, as * ' Laudat Panaetius Africanum, quod fuerit abstinens," *' Panaetius praises Africanus, because he says that he was self- restraining." Fuit for fuerit would mean ** because he actually was self-restraining," without implying that Panaetius said so. So **iniuria quae tibi facta est," the injury which has been done you" ; but iniuria quae tibi facta sit," *' the injury which you say has been done you." Cic. i7t Caec. 58.' ( I ) The translation here marked in italic type I would rather leave to the judgment of scholars than characterise it myself. The correct version is * alleging that he was ' or (better still) * on the ground that he was.^ {2) ' Fuit ' for * fuerit ' would not have been joined by Cicero to such a context as ' laudat quod,' that is to say, where the principal verb is one which by its own nature (as laudo, queror, accuso, &c.) contains Oratio Obliqua, and is used in any person but the first. If the verb has no such nature, as in the well-known passage * Themistocles noctu ambulabat, quod somnum capere non posset,' T. D. iv. 19, Cicero could have written *poterat,' if he had wished to refer the clause to his own statement. (3) Any good scholar, on reading this writer's next citation (from Cic. in Caec. 18) would perceive at once that it is fallacious; that the context, when supplied, must account for the use of *quae sit facta.' And such is the case. Cicero writes : * Hie tu, si laesum te a Verre esse dices, patiar et concedam : si iniuriam tibi factam quereris, defendam et negabo. Deinde de iniuria, quae tibi facta sit, neminem nostrum graviorem vin- dicem esse oportet quam te ipsum, cui facta dicitur.' Then, a few sen- tences later ; * Quid si ne iniiu^iae quidem, quae tihi ab illo facta sit, causa remanet ? ' It would be quite enough to say that for ^ si iniuriam tibi factam quereris ' Cicero might have used the not less frequent ' quereris quod iniuria tibi facta sit,' and that * de iniuria quae tibi facta sit ' is a mere abridgement of Preface. xxix suboblique ' construction : and I suppose this writer has been misled by Madvig, who, in his very faulty treatment of Mood, mentions such examples only. But the principle is general, and applies also to Temporal, Conditional, and Concessive Clauses : as witness the following examples : Darius ejus pontis, dum ipse abesset, custodes reliquit, Nep. Milt 3. At memoria minuitur. Credo, nisi eam exerceas^ aut si sis natura tardior, Cic. C 7. Utilitas efflorescit ex amicitia, etiamsi i\x eam minus secutus sis, Cic. Lael. 27. This last example is gnomic in its nature. See xv. Moreover, it is not true that the Subjunctive, by its own indepejident right, ' denotes an alleged reason or act.^ If this were so, then the compound sentence ' Laudat Africanum Panaetius, nam fuerit abstinens' would be good Latin, and might express * Panaetius praises Africanus, for he was self- denying ' : quod absurdum est, as geometricians say. The truth (overlooked by the writer in question) is that this power belongs to the Mood in subordination only, when it is truly Subjunctive ; and it belongs to it only/;/ its relation to the previous predication, which is never to be left out of question. If such predication is itself subordinate, that is, conveys the thought of another subject going before it, as in (i), then the Subjunctive also shares that thought. If the Subjunctive, as in (2), depends on a principal Indicative (and is not Consecutive, or otherwise controlled), its presence denotes that in that principal predication the idea of Oratio Obliqua is implied. In other words, it is not the dependent mood alone which is then to be considered, but the principal predication together with its dependence. In the sentence cited in the note, ' posset ' con- veys to ^ ambulabat ' the accessary notion of a reason given for the act by Themistocles : ^ poterat ' would confine ' ambulabat ' to the statement of Cicero. XII. § 25. I proceed to support my explanation of this doctrine by citing a large number of examples, which will be more instructive if divided into three classes : * de iniuria, de qua quereris quod tibi facta sit. ' But also * de iniuria, quae tibi facta sit ' is really subordinate to the Oratio Obliqua ' neminem . . . vindicem esse ' : and, when * facta sit ' is afterwards used, Cicero merely cites his own phrase, the import of which is known from the previous con- text. See Example 57, below. XXX Preface, First : Examples in which the text does not exhibit formal oratio obliqua ' \ but a slight addition or a slight change of form at once exhibits it without any difference of sense. Secondly : Examples where ' oratio obliqua ' is implied in the meaning of the principal construction as one of expressed feeling : praise^ b/ame, complaint, accusatio?i, reproach, boasting, giving thanks, promismg, iiidignation, anger, menace, regret, &c. Thirdly : Examples in which no such connexion exists be- tween the principal Sentence and the Clause as to exhibit a manifest ' Oratio Obliqua ' ; but we say, on the ground of analogy, that an accessory meaning is conveyed to the principal construction from the fact of its relation to the Clause. Class I. 1. Ne iustitiam quidem recte quis dixerit per se optabilem, sed ^uia iucunditatis vel plurimum afferat. — Cic. d. Fin, i. 1 6. (Dixerit only wants the dropped esse to make this an example of actual oratio obliqua.) 2. Tefelicemdicis amasque Quod nusquam tibi sit potandum. — Hor. S, ii. 7, 31. {Esse te felicem, ) 3. Recte est igitur surgetque ? nega bit, Quod latus aut renes morbo temptentur acuto. — Hor. S, ii. 3, 162. {^tg2\A\. recte esse,) 4. Hanc reperiebat causa m, quod apud Germanos ea consuetudo esset ut &c. — Caes. B, G, i. 50. {Causam esse.) 5. Cum contemplor animo, reperio quattuor causas, cur senectus misera videatur : unam, quod avocet a rebus gerendis ; alteram, quod corpus faciat infirmius ; tertiam, quod privet omnibus fere voluptatibus ; quar- tam, quod hand procul absit a morte. — Cic. C, M. 5. {Esse being supplied with ' causas,' ' unam ' &c., oratio obliqua exists throughout.) 6. ~ Ille laborem Excusare Philippo et mercennaria vincla Quod non mane domum venisset, denique quod non Providisset eum, — Hor. Ep. i. 7, 66. (Excusare =///V// in causa esse,) Preface, xxxi 7. Bene maiores nostri accubitionem epularem amicorum, quia vitae coniunctionem habe7rt, convivium nomi- narunt. — Cic. C, M, 13. (Nominarunt=m^ ^/jc^rz/;//.) 8. Caesar sua senatusque in Ariovistum beneficia comme- moravit, quod rex appellatus esset a senatu, quod amicus, &c. — Caes. B, G, i. 43. (Commemoravit= multa esse dixit ^ 9. Huic me, quaecumque fuisset, Addixi. — Ytrg.Aeu.iil 652. (Addixi me=^/jt:/ ad- haesurum^ 10. Videor mihi gratum fecisse Siculis, quod eorum iniurias meo labore, inimicitiis, periculo sim persecufus, — Cic. Verr, ii. 6. (Videor m\hx=puto me.) 11. Commodissimum visum est C. Valerium Procillum . . . quod in eo peccandi Germanis causa non esset^ ad eum mittere. — Caes. B. G. i. 47. (Visum tst-=putavit esse.) 12. Mirabile videtur ^//(?^non haruspex cum haru- spicem viderit ; hoc mirabilius quod vos inter vos risum tenere possitis, — Cic. N. D. i. 26. (Mirabile vi- Aetm^mirandum esse putant.) 13. Thucydides libros suos tum scripsisse dicitur, cum a republica remotus atque in exilium pulsus esset. — Cic. d. Or, ii. 15. (Th. scripsisse dicitur = scrip- sisse dicunt.) 14. Quidquid peperisset decreverunt toll ere. — Ter. And. ii. I, 6. {To\[QXQ^-=ut tollerent.) 15. Helvetii constituerunt ea ^^/(7^ ad proficiscendum pej'- tinerent comp Sir siVQ. — Caes. B. G. i. 3, (Comparare= ut compararefit.) Class II. 16. Nemo extulit eum verbis, qui ita dixisset, ut qui adessent intellegerent quid diceret. — Cic. d. Or. \. 14. (Extulit vtxhis-^-laudandum esse dixit ) 17. Athenienses Lacedaemoniorum victorias suae culpae tribuebant, quod Alcibiadem e civitate expulissent, — Nep. Ale. 6. 18. Caesar temeritatem cupiditatemque militum reprehendit, quod sibi ipsi iudicavissent quo procedendum aut quid agendum videretur. — Caes. B, G, vii. 52. xxxii Preface, 19. Nec vero quisquam potest iure reprehendere, quodm?iXQ non transiei'iin, — Cic. Att. viii. 12, 3. 20. Haedui que stum veniebant, quod Harudes, qui nuper in Galliam transportati essent, fines eorum populare7itiu\ — Caes. B, G. i. 37. 21. Saepe illi deplorare solebant, turn qtiod voluptatibus carerent . . . turn qt/od spernerejitiir ab eis, a quibus essent coli soliti. — Cic. 22. Hospitem inclamavit, quods^t^o, absente mihi fidem habere noliiisset. — Plant. Asin, iii. 2.36. 23. Graviter Haeduos a ecu sat quod . . . non sublevetiir ; . . . quod sit dcstitutus^ queritur. — Caes. B. G. i. 16. 24. Theophrastus moriens ac cusasse naturam dicitur, quod hominibus tarn exiguam vitam dcdisset, — Cic. T, D, iii. 28. 25. Vercingetorix proditionis insimulatus est quod castra propius Romanos movisset^ quod cum omni equitatu discessissetj quod sine imperio tantas copias reliquisset, quod eius discessu Romani tanta opportunitate et cele- ritate venissent, — Caes. B, G. vii. 20. 26. Caesar centuriones incusavit, quod aut quarn in partem aut quo consilio ducerentur, sibi quaerendum aut cogi- tandum putafmt. — Caes. B, G. i. 40. 27. Themistocles graviter castigavit Lacedaemonios, quod non virtute sed imbecillitate sociorum potentiam quae- rerent. — lust. ii. 15. 28. Cato obiecit ut probrum M. Nobiliori, ^^/^?^ is in pro- vinciam poetas duxisset. — Cic. T, D. i. 2. ;j9. Litterae ipsae videntur quasi exprobrare quod in ea vita maneam, in qua nihil insit, nisi propagatio miserrimi temporis. — Cic. Fam, vi. i^. 30. Non tam exitu bellorum, quod vincatis, quam principiis, quod non sine causa suscipiatis^ gloriamini. — Liv. xlv. 22. 31. Caesari decima legio per tribunos militum gratias egit, quod de se optimum indicium fecisset. — Caes. (9. i. 41. 32. Themistocles domino navis quis sit aperit, riiulta polli- cens si se conservasset, — Nep. Them, 8. 33. Xerxes ei praemium proposuit, qui invenissei novam voluptatem. — Cic. T. D. v. 7. Preface, Kxxiii 24, Beroen digressa reliqui Aegram, indignantem, tali qu^d sola careret Munere, nec meritos Anchisae inferret honores* — Verg. Aen, V. 650. 35. Augebat iras, quod soli ludaei non cessissent. — Tac. H. V. 10. 36. Atqui voltus erat multa et praeclara minantis^ Si vacuum tepido cepisset villula tecto. — Hor. S. ii. 3^ 9. 37. Aeneas ... minatur Exitium, si quisquam adeat. — Verg. Aeit, xii. 760. See viii. 649. 38. Anpaenitet vo^^quod salvum atque incolumem exer- citum traduxerim% — Caes. C. ii. 32* Class III. 39. Nec fluminibus aggesta terra semper laudabilis, ^2/«z;2^c? senescant sata quaedam aqua. — Plin. N, H. xvii^ 4. (Laudabilis=m quae laudari debeat.) 40. Eo id 1 a u d a b i 1 i u s erat, quod animum eius tanta acer- bitas patria nihil a pietate ave^-tisset. — Liv. vii. 5. (Eo laudabilius erat=. v. 36 (given by us, p. 459) Aristides, &c. and the two following : Fabio dicta dies est, quod legatus in Gallos fugnasset^ Liv. vi. I. Aedem lovi vovit, si eo die hostes fudisset^ Li v. xxxi. 21 : (in which obviously : Fabio dicta dies tsi=Fabius accusatus est^ and vovit contains se dedicaturum). Thus, by saying * without a formal use of the Obliqua Oratio,' Key recognises an informal (or virtual) use of it, as I do ; and postpones this rightly to the formal use. I could cite German grammarians, were it worth while, whose treat- ment implies the same principles : for instance, Middendorf and Griiter, Frei, Billroth, Ellendt, &c. But the term (Virtual O. O.) was, I repeat, introduced by me thirty-six years ago. xl Preface. It rests upon its own fitness : I can but deprecate, if it exist anywhere, the spirit complained of by Horace, when he sa ys Indignor quicquam reprehendi, non quia crasse Compositum illepideve putetur, sed quia nuper. Ep, ii. I, 76. XVIII. § 31. The question, whether the (independent) Thought-mood should be called Subjunctive or Conjunctive, stands as follows: The Greek grammarians of Alexandria used the term tyKkidiq vTroraKTiKYj, modus subjunctivus. Why? Because in Greek there are two forms of the Thought-mood, one of which they called evKTiKrj, Optative, the other vworaKTLd], Subjunctive. Neither of these terms corresponds exactly to the uses of the respective orms. The term Optative expresses only one use of the first: — that of praying or wishing, eXdoi, may it come: but it has also a dependent use, on eXdoi, that it was come) and by the convenient accession of the modal particle av it gains an in- dependent or enuntiative power tXdoi av, like the Latin Weni- ret,' // would come. The second form (Xdri was called viroraK- T Ki] because it never did acquire enuntiative power ; the modal av was not extended to it, but only the conjunction ar, if, the conjunctional relative og ay &c,, whosoever, &c., in dependent construction. It has, however, an independent power as suc- cursal to the imperative, in hortative sense ist pers. plur., eXdiojjtev, let us come ; and as interrogative, in dubitative sense, — 71 aj, what can I say ? In spite of these two exceptional uses, it is manifest that the term vTroraKTiKy, subjunctive, is, for the Greek mood, fully defensible, because its principal and (so to say) normal use is dependence, German grammarians, however, call it Conjunctive ; wisely we think, for the maintenance of analogy But lor calling the Latin Thought-mood, generally. Sub- junctive, there seems to be, from a right point of view, no reason- able defence. Key, indeed, has taken a point of view, which, if it were right, would supply one. His words are (Gr. §§ 427-8): *The Subjunctive Mood, as its name implies, is used in secondary sentences subjoined to the main verb. In some sentences it is not uncommon to omit the main verb, and then the Subjunctive Mood seems to signify power^ permission^ duty, Preface. xli wishy purpose, result, allegatio7i, hypothesis \ whereas in fact these notions belong to the verb which is not expressed. Thus the phrase " quid faciam " is translated by what should I do or what am I to do ? but the full phrase is " quid vis faciam ? " what do you wish me to do? (!) ' This theory Key, perhaps, borrowed from Hermann, who applies it (De emendanda ratione Grammaticae Graecae) to explain the two exceptional uses before noticed of the Greek Subjunctive: supposing 'iojfxey = aye 'iiofxev, and ri in the central districts, and the Latin in Latium. Umbrian, Sabellian, Oscan, and others were destined to fade away, leaving a few scat- tered monuments of their former existence. Latin survived to be the parent of learning and language in Western Europe. Rome, founded on the Tiber by Latins, according to tradition, B.C. 754, became, on the fall of Alba, the head of the Latin race and name (nomen Latinum) ; and the clannish pride of the Romans led them to call their language, and afterwards their literature, Latin rather than Roman. 3. By Roman conquest and dominion the Latin speech was ex- tended, with dialectic varieties, to all Italy and to other neighbour- ing countries. From this source are derived the following modern languages : Italian, French (in both its divisions, Oc and 0/7), Spanish, Portuguese, Wallachian, and the Romansch of the Swiss Grisons. They bear the common title of Romanic or Romance languages. All are more or less alloyed with the Teutonic dialects which barbarian conquest carried into Western and Southern Europe in the fifth and following centuries. 3 English. English is the single instance of a Teutonic language largely alloyed, without being disorganised, by the speech of Romanic con- querors. When the Romans quitted Britain in the fifth century, the island, after a brief interval, was overrun by Teutonic hordes (Saxons, Angles, and Jutes), who formed no fusion with the Keltic natives, but either extirpated them gradually, or drove them (as Walsche, Welsh, or foreigners) into mountainous and barren dis- tricts. The rest of the country south of the Tweed came to be called England (Angle-land), and its speech (Anglo-Saxon) was the parent of the later English. The conversion of the Saxons to the Christian faith brought into England some knowledge of Latin, and incorporated many Latin words with the English tongue. By the Norman conquest, A.D. 1066, a dominant race came in, who, though comparatively few in number, filled most places of rank, power, and influence. Hence their speech — Norman-French, a Romanic dialect — became that of courtly society and of law; Latin, its mother-tongue, became the vehicle of religious service and learned intercourse ; whilst English continued to be spoken by the great bulk of the population. In the fusion of these varieties, by which modern English was gradually formed, the usage of the yeomanry and peasantry prevailed over that of the nobles, the law, and the church. English is structurally a Teutonic language, and the number of Teutonic words holds to those of Latin origin a propor- tion of about two to one. This shews that, without a knowledge of Latin, it is impossible to gain a thorough knowledge of English. It must also be remembered that the Teutonic element in English has itself a distant kinship to Latin. Influ- The influence of Greek civilisation upon Latin was immense. 2nce of Besides their original affinity the Greek race came into influential contact with the Latin at two distinct eras. The first of these was §5. Introduction. when the Greek colonies in Sicily and Italy became active in com- merce and literature. This activity may be dated as beginning about 550 B.C. The Aeolic city of Cumae in Campania appears to have been the chief medium of communication between Rome and the Greek colonies, and to the influence then exercised may perhaps be ascribed those facts of language which led grammarians to derive Latin from the Aeolic Greek Dialect. Hence too the Romans probably drew the peculiarities which characterise the Latin Alpha- bet, as the letter Q and the V consonant, which the Aeolic Greeks had kept in the Dorian alphabet at Cumae. 5 Again, when literary activity began at Rome in the third Sketch century B.C., Grecian literature supplied most of the forms and Lkera-^" much of the matter. Rome had no models to furnish. Inscriptions, ture. laws, crude annals, with fragments of ritual songs and coarse farces, are all it has to shew within its first five centuries. The credit of authorship is ascribed first to Livius Andronicus, who wrote dramas for the stage B.C. 240. He was succeeded by a crowd of authors, among whom may be mentioned Naevius, En- nius, the father of epic poetry at Rome, and Lucilius, whose subject and reputed invention, satire, is the most original product in Latin literature. But of these writers mere fragments remain. The comedies of Plautus (Plant.) ^ and Terentius (Ten), founded on those of the later Attic stage, with the remnant De Re Rustica of the elder Cato, are the only literary works extant in Latin before 85 B.C., the date of Cicero's earliest writings. From this time to A.D. 14 extends what is usually called the Golden Age of Latin. Its most eminent authors are : — Prose. Cicero C. (or Cic.) Caesar Caes. Cornelius N epos... N. (or Nep.) Sallustius Sail. Livius L. (or Liv.) Varro Varr. Vitruvius Vitr. Poetry. Lucretius Lucr. Catullus Cat. Vergilius V. (or Verg.) Horatius IJ* (or Hor.) Tibullus Til>. Propertius Prop. Ovidius Ov. The so-called Silver Age, to a.d. 117, contains among others : Prose Seneca Sen. Quintilianus Qu. Pliniusthe elder... PI. N. H. Plinius the younger Plin. Valerius Maximus V. Max. Velleius Paterculus Veil. Tacitus Tac. Suetonius Suet. Florus ? Fl. Q. Curtius? Curt. Poetry. Manilius Man. Phaedrus Phaed. Seneca Sen. Tr. Lucanus Lucan. Persius Pers. Silius Italicus S. It. Valerius Flaccus V. Fl. Statins St. luvenalis luv. Martialis Mart. * The letters following the names shew the abbreviations used for them in this Grammar. B 2 Introduction. §5. The next period, extending to the fall of the Western Empire, A.D. 476, has been termed the Brazen Age. The writers who come nearest to the classic style during this period, are : — Prose. A. Gellius Gell. lustinus lust. Appuleius App. Eutropius Eutr. Macrobius Macr. In the Iron Age, w^hich succeeded, Boethius may be named as the most successful imitator of classic purity.* Poetry. Ausonius Aus. Claudianus Claud. * Other abbreviations used in this Grammar : Pr. Primitive (Sound or Root). Sk. Sanskrit. Gr. Greek. E. L. Early Latin (before 186 B.C.). R. L. Republican Latin (from 186 to 30 B.C.) I. L. Latin of Imperial Age (from 30 B.C. to 170 A.D.). C. L. Classical Latin. L. L. Later Latin. U. Umbrian. O. Oscan. S. SabelHan. F. Faliscan. V. Volscian. M. Lucr. Munro on Lucretius. C. Corssen (Aussprache). , Curt. G. Curtius (Gr. Etymologie). Three dots (...) following a word imply that other derived or kindred words are to be included. In Sanskrit words : represents the palatal sound ch (as in * church'): ric' is sounded 'rich.' G. Cur- tius represents it by . s' represents the slightly aspirated sibilant, which often corresponds to Greek k and Latin c, q. Sk. c^as'an, Gr. 6e/ca, L. decem. Sk. «TzV, Gr. deiK-, L. doc-eo, 8lc. G. Curtius represents it by c. rt is a Sanskrit vowel, which may be written ar. See p. 578. y is the Sanskrit letter= English j (Curtius^')* English y-consonant (Curtius j). Ex.— yuj, to yoke (Curt. jug'). PART I. LATIN ETYMOLOGY. Etymology comprises : — Etymo- logy— L Phonology or Soundlore, the doctrine of ^l^^^^^' Sounds. 11. Morphology or Wordlore, the doctrine of Words. ^ By a Primitive Sound or Root is meant one which careful in^ duction assigns to that ancient, though no longer extant, Aryan language from which the Sanskrit is derived. Such induction is obtained by comparison of the Sanskrit with all other kindred languages, especially with Zand, Greek, Latin, Gothic, and Lithu- anian. See Supplementary Notes following Appendix. DIVISION I. PHONOLOGY OR SOUNDLORE. 7 i. Soundlore treats of the sounds and relations of Sound- Letters and Syllables. *°*^^* 1. The Latin Alphabet now in use contains the Latin same Letters as the English, omitting W. ^^^^* The Letters have two forms : 1) The Capital, Uncial, or ancient form.. 2) The Small, or later form, which came into commoa use in the eighth or ninth century: after which the Capitals were chiefly used for inscriptions, and as initial letters of sentences and proper names. 1) ABCDEFGHI(J) KLMNOPQRST 2) abcdefghi(j)klmnopqrst (U) V X Y Z. (v) u X y z. 2. Six of the Letters are VocALESy Vowels (self-sound- ing), a, e, i, o, u, y : the rest are ConsonanteSj Con- sonants, which are sounded only with a vowel. ' The terms Phonology and Morphology are taken from Schleicher's Vergleichende Grammatik der Jndogermanischert Sprachen. Latin Soimdlore. §8-9. Sylla- bles : Diph- thongs. 8 (^)iian- tit V. 9 Syllaba- 3. Consonants are divided into Mutes, Nasals, Liquids, Spirants, and Double Consonants.* The Nasals are n, m ; Liquids, r, 1 ; Spirants, f, h, j, s, V ; Double Consonants, x, z : the rest are Mutes. Note I. y and z are only used in words borrowed from the Greek. Note 2. l+j and u + v are two pairs ; each pair constituting one ancient letter in double form. See § 12. viii. 4. A Syllable {avWa^r}) consists of one or more letters pronounced in a single breath ; i-lex. 5. A Diphthong (8/^^0770^) is the combined sound of two vowels meeting in the same syllable ; au-lae. There are in Latin three usual diphthongs, ae (or ae), oe (or oe), au ; and three seldom used, ei, eu, ui. ii. Quantity \s thd X.\m^ oi uttering a Syllable. I. Every Syllable is considered Short {J) or Long (") in Quantity, according as its vowel is short or long ; that is, accord- ing as it is uttered with a single .or double time (mora) : / Short by nature . . a 6 in . . amor. Long by nature . . e u in . • esu. Short by position before another vowel . . It in . . . pTos. Long by position before two consonants or a L double consonant A Vowel may be •< e o m . au, ae in pernox. caudae. Diphthongs are long .... 2. A Vowel is called Short or Long by Nature, when the reason of its quantity is other than position. 3. A Syllable is called Doubtful (-) when its Vowel may be short or long : Sidonius. 4. A Vowel, naturally short, may be made long in poetry, if it stands before cr, gr, tr, dr, pr, br, fr, cl, pi, or fl : tenebrae, quadrSplex. Such a Vowel is called Doubtful by position. Id prose the syllable is pronounced short, tenebrae. iii. Syllabation is subject to the following rules: 1. Every syllable must contain a vowel. 2. A word may begin with any vowel but y. 3. A word may end with any vowel, and with any of the con- sonants, 1, m, n, r, s, t, x. A few words end in b, c, d. • Checks are another term for Mutes; Trills for Liquids; Fricatives (as gene- rated by the friction of the breath) for Spirants (See Max yinWer's Lectures on Langtiage). Mutes have also been called Momentaneous or Explosive Consonants, as distin- guished from Nasals, Liquids, and Spirants, which are Continuous. Accentuation. 7 4. Priscian's rule is that inner syllables end with a vowel, if a single consonant follows : cla-ma-tur ; or if two or more conso- nants follow, which can begin a word : lu-di-crus, e-sca, ma-gnus, scri-ptus, scri-psi, a-stra. Custom extends this rule to such in- stances as so-mnus, A-bdera, rhy-thmus, etc. But, if the conso- nants cannot begin a word, they are divided between the syllables : gal-lus, punc-tum, of-fen-do, am-plis-simus, ex-per-tus. 5. In compound words a syllable ends with the end of one part : ab-igo, res-publica. 6. The last syllable is called Ultima, the last but one Penul- tima or Penult, the last but two Antepenultimaor Antepenult. An inner syllable is called open if it ends with a vowel, close if it ends with a consonant. A word of one syllable is called Mono» syllable; a word of two syllables, Di syllable, etc. vi. Accentuation h^s ihe ioWovjingldcVJsi^ Accent. uation. 1. The tone of a syllable is called Accent. There are two strong accents, the Acute (') or sharp quick stress, and the Cir- cumflex (^) or deep lengthened stress. Syllables without either are sometimes called Baryton (Grave in tone), sometimes Atonic. 2. Monosyllables, with vowel short by nature, have the Acute Accent : 6s, vir, dux; those with Vowel long by nature have the Circumflex : 6s, mo s, lex. 3. Words of several syllables are not accented on the Ultima, but on one of the two preceding syllables. Only, when a word loses a final vowel, if the Penult had an accent, that accent remains on the same syllable : illinc for ilHmce, audin for audisne. 4. Disyllables have the Acute on the Penult, when either both syllables are short by nature, as b6na, or the first is long by posi- tion, or the last long by nature or position, as inter, m6res, amant. 5. Disyllables have the Circumflex on the Penult, when the Penult is long by nature, and the last short by nature and not lengthened by position :mater, musa. 6. Words of more than two syllables have the Acute on the Antepenult, when the Penult is short, as hdmines, amph'ssimos. But the accent remains on the Penult in unweakened compounds pf facio, as benef^cis, and in contracted genitives : ingeni. 7. Words of more than two syllables have their Accent on the Penult when long ; namely, the Acute when the Penult is long by position only, clamantur; or when the last syllable also is long, am a rant ; the Circumflex, when the Penult is long by nature, and the last short by nature, and not lengthened by position : clamare. 8. Enclitics, que, ne, ve, etc. bring forward the accent of the word to which they are subjoined : homines, homines que; pr6na, prondque. Prepositions are Proclitic, that is, with- out an accent of their own before their cases . circa moenia; but, moenia circa. Note, The Accents are not marked on words, but understood.^ ' Latin Accentuation is a subject too large, intricate and unsettled to be fully discussed in a book like the present. Students who wish to pursue it more minutely are referred to the works of Ritschl and Corssen, and to the Latin Grammars of Kriiger and Kiihner. Only a few general rules are given here. * Corssen allows a second accent, which he calls a Middle Tone (i.e. semiacute), to 8 Latin Sotmdlore, § II-I2 V. PuNCTA, the Signs of Punctuation, or Stops. These are the same in Latin as in Enghsh : Comma (,) ; Semi- colon (;); Colon (:); Full Stop (.) ; Note of Interrogation (?) ; Note of Admiration (!). The mark ( •• ) is placed over e or l when it does not coalesce with a preceding vowel: aer, Teius. But none of these were used anciently except the Full Stop (punctum). vi. The Relations of the Letters may be thus shewn : I. VOWELS. Standard vowel a Sharp medial e o Flat medial Sharp semiconsonant 1 |- u Flat semiconsonant Medial (see vii. ii). II. CONSONANTS. Note. — T e n u e s and Mediae are by some called severally Surds and Sonants. Guttural mutes are also called k-sounds, Dental mutes t- sounds, Labial mutes p- sounds. Mutes i (Momentan- eous) ! >- Nasals Liquids (Continu -^ Spirants ous) ' Double Tenues or Sharp s f Sibilant i) ■«-> rt >-i 'a, < Semivowel Guttural ^oxT\ixo2X- \ sounds . . . C (1^, q) • > • • X Dental, or Teeth- \ sounds . . . i t a n 1 S • • f j V 1 z Labial, or Lip-sounds p b m 1 j N is guttural when it precedes gr, c, or q. The Spirants f, v, if sounded as in English, are labiodental.^ certain words of more than three syllables, when there is an interval between the ac- cented syllables. Such words are — 1) Plurisyllable Compounds, in which the Middle Tone will fall generally on the first syllable, as in versipellis, misericordia, undeviginti, efifrenatus, sublevare, etc. ; sometimes on the second, as in superbiloquentia, repandirostrum : 2) Plurisyllables, in which, by Derivation or Flexion, the accent of the primitive word has been shifted to a suffix. Such a Middle Tone will generally be on the first syllable ; as in Idngitudo, pdpulabundus, servitutem. This theory implies, in rare instances, the possibility of two Middle Tones, as in cdnfidentiloquius. See Corssen, ii. 824. * Sanskrit has two more classes of Consonants : (a) Palatal, a modification of the Gutturals ; (b) Cerebral or Lingual, a modification of the Dentals. Thus, in Sanskrit J is Palatal. Semiconsonants. 9 vii. Memoranda from the History of the Al- History u 4- of Al- pha bet phabet. 1. The Romans modified the form of the third Greek letter from r to C, and gave it the sound K, instead of G. The sound and letter G were afterwards introduced about 250 B.C. C. was kept as the abbreviation of Gains ; Cn. of Gnaeus. 2. The use of C as sharp made K superfluous, and the Romans almost ceased to use it ; but it was kept in a few abbreviations : K. for Kaeso; Kal. for Calendae and Calumnia: also Kar. for Kartago. 3. The other Phoenician guttural surd Koph or Koppa (which the Dorian Alphabet of Cumae possessed) was kept by the Romans as Q, and ultimately confined to words in which parasitic u or v follows the guttural. 4. The Rough Breathing was raised to the rank of a letter, ob- taining the form and position of the Greek Eta, H. 5. The letter f was purely Italian, its ancient sound being not exactly that (which it now has) of Phi (ph), nor that (which its form suggests) of the Greek Digamma (w). In the Etruscan Al- phabet it has the form 8 ; but the Romans gave it that of the Di- gamma, (f) F. 6. Vau (V nearly = English w) was adopted by the Romans as a semi-vowel, and took the position held by Upsilon next to T. 7. There is some evidence that the ancient Romans used z (zeta), but afterwards supplied it by s or ss : so that y, z were added at the end of the Alphabet in Cicero's age to represent the Greek v, and are only used in latinized Greek words ; lyra = \vpaj zona 8. The history of x is obscure. It appears in a few early in- scriptions : but, though in power equal to Greek £ (cs), it took the place and form of Chi : when and why, are doubtful points. See Corssen {Ausspr. I. 6). 9. The long vowels Eta and Omega were omitted as unnecessary. 10. The Aspirate sounds x> ^1 0? as foreign to Italian utterance, were left out ; but the study of Greek in Cicero's age led to the use of ch, th, ph, which represent those letters in latinized Greek words : parochus, thesaurus, philosophus ; also of rh : rhetor. 11. The Emperor Claudius invented and introduced three letters : (i) j to represent u-consonant; (2) 3 (antisigma) to represent ^ (ps) ; (3) h to represent a vowel having a middle tone between 1 and u, as in libet — liibet, gradibus — gradubus, maximus — maxumus. They did not remain in use ; but the first and last appear in inscriptions. viii. The Semiconsonants 1 (j) and u (v). I. The consonantal character of 1 (j) is shewn by the two facts, I (J), that, when it begins Latin words before a vowel it makes position in verse after words ending with a consonant, and that it makes no hiatus after words ending with a vowel or with m. Thus in Sub loue iam ius est sub, iam are long by position, and -e, -am are not elided before t. 10 Latifi Soimdlore, 1) 1 is a Vowel (l-vocalis) when it ends a syllable ; ni-ti ; or when it stands in a syllable before a consonant, in-it, sic-cis; or when it is a syllable ; ab-i-to. I is a consonant (l-consonans) when it begins a syllable before a vowel ; i e-i u-n o ; its sound being that of English y-consonant ( = German j), a faint protraction of^the vowel-sound 1. 2) In Greek words, however, 161 e, iambus, and in a few Latin words, i-ens, ieram, i-vocalis remains open before a vowel. 3) The sign J was introduced in a late age, to represent l-con- sonans, and most editors do not use it. Its English and French sibilant sounds {John^ Jean) are not classical, but crept in before the Middle Ages. 4) l-consonans is omitted before 1 by the compounds of lacio, ab-icio con-icio pro-icio re-icio, etc. ; though the long quan- tity of the first syllable is kept ; i being =j i. See Munro on Lucr, i. 34, ii. 951. In Lucr. Verg. eice, reice. But re-icere. Plant. 5) Poets sometimes harden l-vocalis into l-consonans : ab-ie-te, ar-ie-te, par-ie-te, for ab-i-ete, etc. Sound ab-yete, ar-yete, etc., trisyll. So Horace has consil-ium ( = consil-yum), Virgil has fluv-iorum ( = fluv-yorum), trisyll. M. Lucr. ii. 991. V (u). 2. The sign V was employed by the Romans as vowel and con- sonant. In a latter age u became the vowel sign, v the consonant sign. If uva ( = oowa) be sounded, it appears that u-consonans ( = w) is only a faint protraction of the labial vowel u ; whence the modern name Double-u. 1) V-consonans is vocalized in cautum for cavitum, fautum for favitum, lautum for lavitum, and in auceps for aviceps, nauta for navita, naufragus for navifragus. 2) Poets sometimes vocalize u-consonans before a vowel: sil-u- ae : sometimes they harden u-vocaHs into u-consonans : gen-ua for ge-nu-a, ten-ui-a for te-nu-ia. M. Lucr. iv. 1157. 3) Parasitic u follows q, ng:, and s : sequor; lingua; suavis. This usage is derived from ancient groups kv, gv, sv : but as the sign u so used neither forms a syllable nor creates position, it must be regarded, not as a proper letter, but as a kind of link between the guttural (or sibilant) and labial sounds.^ ix. Sound and Quality of the Vowels. Sound I. Vowels have not one short and one long sound only ; but ^aiit various shades of these, in close or open syllables, of ^ ^ (Thus the sound of u varies in the following words : credulus^ Vowels, busy^fiilly use, yune, and in the French words, commiin^ com7nune.) 2. The old sounds of the Latin vowels probably differed little from those of the vowels in modern Italian. Proceeding from the thinnest and sharpest sound i, to the thickest and flattest u, the following words may represent their general distinction : the first four being pronounced as in French, the fifth as in Italian, Zuloo. Quinine, demesne, papa, promote, Zulu. 3. The three primitive vowels are a, i, u. Sanskrit has e and o only as diphthongs arising from al, au. * In many modern editions of Latin authors, V alone is used as the Capital form of con- sonant and vowel, and U alone as the Cursive form of both. In this grammar v is retained as a cursive. Phonetic Decay, II 1) The standard vowel is a, issuing from the throat through the opened mouth : 1 is the thin sharp palatal, sounded between the tongue and the lower palate ; u is the thick flat labial, sounded by a low interior whistle through the protruded lips. Each has its long and short sound, with shades of these. 2) The want of intervening sounds to represent the strengthen- ing of 1 and ii, and the primary weakenings of a, called into use two subsidiary vowels ; e medial between a and 1, and 6 medial between a and u. Both these are narrower gutturals than a ; e sounding along the upper palate and tending to the sharpness of i ; and o sounding from the lower throat with a fullness which its form marks, but tending to the labialism and flatness of u.^ 3) The strongest short vowel is a, into which none other passes. 4) The weakest is 1 : for which reason it often stands as a vin- cular vowel before suffixes : reg-/-to, fiag-/-to, leg-^-bus : but sometimes e or u takes its place; soci-/-tas, teg-//-mentum. 5) That 6 is stronger than e may be seen by comparing pondus with pendere, toga with tegere, volo with velim, velle. 6) Though u has various shades of strengh, as in put o, siimus, augur, augurium, declining almost to the weakness of 1, as in optumus (optimus, vii. 11), yet on the average it is not seen to be weaker than e. Such examples as pignus, pigneris, pignoris etc., might seem to shew ii stronger than e, 6 : but it must be re- membered that this u corresponds not to Greek v, but to Greek o : that us is really a weak syllable, and u, like 1, gives a facility to the rejection of s in old Latin poetry, which e does not afford. 7) When a vowel from being short becomes long, it is doubled in time and strength ; X 1 = i, ee = e, etc. 8) As final short vowels, a and e predominate ; 1, 6 are rarely final ; vl never, except by the rejection of s in old Latin poetry. As final long vowels, i and 0 predominate*: e and a are less frequent ; and u least frequent of all. X. Phonetic Decay in old Italian language. Pho- (See Corssen, L 347.) D ^cay I. Phonetic Decay tends to lighten diphthongs, to shorten and weaken vowels, to silence or throw out light vowels, to cast off" or assimilate consonants. . 2. This tendency is especially shewn in Umbrian and its cognate dialects ; also in the old Latin, of which our knowledge is derived chiefly from inscriptions, partly from the testimonies of gramma- rians, and from the most ancient manuscripts. 3. Classical Latin (see § 5) may be regarded as in some measure a reaction, by which, during a long literary period, the process of Phonetic Decay in Latin was arrested. After the age of Sueto- nius, about A.D. 120, decay recommenced and continued for nearly 1000 years, till the modern Romanic languages gradually emerged from the darkness of those centuries. ^ Vowels are here spoken of as sounded with consonants : ad, da. All vowels issue from one primary vowel, a faint sigh in the depths of the throat. Its first scarce distm- guishable strengthenings may be compared thus : max-t-mos, raax-w-mos, max-^-mos* max-J-mos, max-a-mos. 12 Latin Soundlore. §12. Vowel- Xi. VOWELCHANGE. 1. Vowels are liable to change in the Flexion, Derivation, and Composition of words. 2. Syllables may be either strengthened or weakened by Vowelchange. 3. The general tendency of Italian dialects was to weaken vowel sounds. But sometimes a syllable is strengthened by assuming a stronger for a weaker vowel. Thus the tonic syllable in toga is stronger than in teg o. Diph- xii. Formation and Decay of Diphthongs. thongs. 1. Diphthongs and long Vowels in Sanskrit arise from the intro- duction of a vowel to strengthen a short sound. Thus a by strengthening 1 produces e ; a by strengthening ii produces 6 ; and this process is called Guna (distinction). If a strengthens (ai, au), the process is called Vriddhi (augmentation). 2. In Latin (as in Greek) e and 6 are themselves capable of strengthening 1 and u. Thus the full list of diphthongs proper is al el ol au eu ou 3. ui is an improper diphthong, only found in the words cui, huic, hui, phui. 4. Yi in Greek words for vl is very rare : as Hi thy i a (£A\uQv1a). 5. The six diphthongs proper existed in ancient Latin, as shewn by inscriptions : but before the classic age all except au had de- cayed into other long sounds, namely : — ai into ae, rarely passing into e (el) i ol — oe, often — — u (el) i el — e or i eu — u ou — u Note. Corssen observes (I. 674), that the history of al, ol, el, has peculiar interest, because, as these diphthongs often sprang from the addition of a suffix which begins with a vowel to a stem which ends with a vowel, they illustrate the laws of flexion as well as the progress of phonetic decay.^ 6. The diphthong al prevailed in old Latin: aidilis, Romai, filiai, &c. ; and is found even in the imperial age. But about B.C. 200 ae came into use, and gradually became the classical form. For this, as a rustic variety, in the age of Lucilius, is found e : edus, pretor, Cecilius, &c. ; which became more prevalent in later Latin, and in modern language has superseded ae : secular, premium, i) Examples occur of els for Dat. and Abl. Plur. Ending als, which in classical Latin became is : tabuleis publiceis = tabulis publicis. * This history is gained from the careful comparison of Latin Inscriptions extending for about 400 years from B.C. 260 to a.d. 150. In the present chapter and in § 20 such results alone are generally mentioned as suffice to explain the varieties of form which appear in classical authors. Formation and Decay of Diphthongs. 13 a) Old poets, as Lucretius, often use the Gen. in a-i, dividing it into two syllables, materia-!. Virgil rarely : aula-i, Aen. iii. 354. b) The vowels are divided in Ga-i-us, Ga-T, and in the Greek words Agla-i-a, La-i-us. In others, as Achaia, Aiax, Grains, Maia, and in aio, maior, i is i-cons. (Acha-ya, a-yo, ma- yor, .,.). 7. The diphthong ol (oe) is of much rarer use than ai (ae). Oi is found as late as the first century B.C. in stem-syllables. But it passed into oe early, as ai into ae, by strengthening i. Also, by weakening o, it passed into (ui, ue = ) u. Thus we find moi- nera, moenera, miinera; loidos, loedus, ludus; Troirrjy poena, punire; coirare, coerare, curare (also courare) ; oitier, oetier, iiti. M. Lucr. ii. 829. 1) In the Imperial age oe began to be corrupted into (ee) e : pomerium, federatus. This, as in ae, grew more and more usual, and prevails in modern language : penal^ federal^ &c. 2) In Case-endings, ol from old times was liable to pass into ei andl: puer-oi, puer-ei, puer-i ; puer-ois, puer-eis, puer-is. a) In proln, proinde, the vowels coalesce : in Tro-i-us they remain separate (Hiatus). In Troia I is i-cons. (Tro-ya). 8. Eiis either a diphthong, as in a few old words, deiva, deicere^ leiber ; in the old Italian Dative ending ei, virtutei ; in Dat. Abl. PI. endings eis for ois ; and in Nom. PI. ending ei for oi of Decl. 2 : or it is a middle sound between e and i, as when the form eis represents a compromise between the Acc. Plur. endings is, es, of I-nouns : urbis, urbes, urbeis. i) Ei is found as diphthong or middle sound in inscriptions of all ages ; e sometimes taking its place, but i prevailing over both. In the words dein, deinde, deinceps, e-i may coalesce in poetry, or, as some think, e is elided. b) In names in -eius, i is i-cons. ; Pompeius = Pompe-yus. c) Greek et appears in Latin generally as i: Tydides for Tvhdhi]^\ but often as e before a vowel ; Alexandrea or Alex- andria, Dareus or Darius, Thalea or Thalia. d) In some Greek words e is open before i (Hiatus) : Teius, Plei'as, Nereides. » 9. Au was retained in all ages of Latin. But it often passed (by the process ou, 00) into 6 : codex, plostrum, Clodius, Plotius, &c., and (in Italian) oro, toro, iesoro, &c. ; sometimes (by the process ou, uu) into ii : cliido. M. Lucr. ii. 829. 10. Eu in Latin words is very rare. Heu, eheu, alas, are imitative words. In neu, seu, ceu, u is a vocalised v (neve, seve, ceve). On neuter, neutiquam, see Prosody. d) In Greek words eu remains : Euripides, Eurus. Note, — Greek au and eu before a vowel are written with vowel or consonant : Agaue or Agave ; Euander or Evander. 11. Ou occurs on old Inscrr. : doucere, iotiszt, Louceria\ but had decayed into ii before the classical age. (On the pronunciation of Diphthongs, see Appendix B,) 14 Latin Soundlore, Vowel- xiii. VowELSTRENGTHENiNG in Root-syllablcs, onTnT^' Suffixes, and Endings. A) Root-syllables (see Supplemental Note6 after Appendix). d) Primitive or Italian n, strengthened by a (6, e) in the manner of Guna, or by a, (5, e) in the manner of Vriddhi, subsides from a diphthongal sound into u or d ; the series of possible change being u, au, ou, eu, u (d). jug, to yoke kru, hear . kru, be raw lu, wash . plu,^^?7£/ . (pu-s, nur- iiire^ thrive) pu, cleanse pii, stink . ru, bray rudh, be red ruk, give light skii, hide su, sew (u) iu(n)g-ere ... iug-u-m, iug-are (eu) ^fvy- (u) iug-i-s, iug-er-a, iu-mentu-m. (ii) clu-ere. (au) lau-s, laud-are .... (u, o), lu- s-cin-ia, glo-r-ia (a) ru-di-s. (au) rau-du-s, rau-du-s-culu-m. (u) cru-du-s ... cru-d-eli-s ... cru-s-tu-m, cru-s-ta. (u) lii-ere ... lu-tu-m .... (au) lav-ere with v- cons. ... lautus. (ou) \ov-ctr, di-luv-iu-m ... with v-cons. for v-voc. (u, o), l6-tu-s ... pol-lu-tu-s ... lu-s-tru-m, lu-s-tr-are .... (ii) plu-it. (ou) pluv-ia ... for E. L. plov-ia ... (u, o), plu-vi ... pl6-r-are .... (u) pu-er ... pii-ella ... pu-s-illu-s ... disci-pu-lu-s, pu-m-ilu-s, pu-tu-s. (au) ttqIc for ira-vg. (u) pQ-su-s, pu-s-io, pu-pu-s, pii-pa, pu-p- ula, pii-p-illu-s, pu-m-ilio, pu-ber, pu-bes. (ii) pu-tu-s, pu-t-are ... am-pii-t-are. (au) pa-eni- t-et for pav-ine-t-et. (ou) po-ena for pov- ina. (ii) pii-ru-s ... pu-n-ire (ii) pu-ter pu-tris .... (au) pa-e-d-or ior pav-i-d-or, (u) pu-s pii-r- ... pu-t-ere .... (ii) ru-d-ere ... ru-d-en-s. (au) rav-u-s ... rau- cu-s .... (u) ru-m-or. (ii) rub-er, rub-ere ... ru-tilu-s ... epvB-pog. (eu) ip£vd-(o, (u, o) rub-ig-o or rob-ig-o, ruf- u-s .... (u) luc-er-na, \vk-. (ou) E. L. Louc-ina. (eu) XevKog, (vl) lux luc- ... luc-ere ... Liic-ina, di-luc-ulu-m .... (u) scu-tu-la ... cu-ti-s. (au) cav-ere, cau-tu-s ... cau-s-sa or cau-sa. (ou) E. L. cou-r-are or coi-r-are for cov-i-r-are. (eu) Kev-6w» (ii) cu-ra, cu-r-are ... E. L. cd-r-are, cu- s-t-o-s, scu-tu-m, ob-scu-ru-s. See C. I. 356. (u) su-ere. (diphthongal in Sk. Goth, and Lith.). (ii)su-t-or, su-bula. § 12. Vowclstrengthening. 15 tu, swell . (u) tu-m-ere, tu-m-idu-s, tu-mu-lu-s. (au in Sk.), tae-d-et for tav-i-d-et. So ta-e-ter. O. tau-ta. (ou) O. tou-to. (u, 5) t5-tu-s, tu- ber. U. tU'ta, to-ta. These old Italian words mean ^ a cotnmunityy a people,^ (u) dux due- ... educ-are .... (ou) E. L. done- * ere. (u) duc-ere .... (u) flu-ere ... con-flu-g-es. (ou) fluv-idu-s, fliiv- iu-s (from flou-v-). (u) flu-men, flu-t-are. (ii) nu-ere. (eu) vevit), (u) nu-tu-s, nu-t-are ... nu-men. Add the prim, root dyu, shine ( = div), whence (6 for ii) io-cu-s. (ou), O. Diouv-ei, whence the weakened words, E. L. Didv-is, U. Juv-e, I6v-is, iuv-at, .... (u) lu-p-piter, lu-n-o, iu-c-undu-s.... ' b) Primitive or Italian 1, strengthened by a (6, e), or by a (o, e), subsides from a diphthongal sound into i or S ; the series being 1, ai, oi, ei, i (e). due-, lead ^n-yjlow nu-, nod dtk, shew. div, shine that k!, lie down smear . spTc( = spak), spy trT, three . fid, to trust lib, to choose (1) dic-are ... dic-io, con-dic-io, dic-is, caussi- dic-u-s ... iu-dex, iu-dic- ... di-dic-i, dig- i-tu-s, dig-nu-s ... pro-dig-iu-m ; ... IXk-y) ... . (ei) ^eiK'E.L. deic-ere. (i) dic-ere, dix-i ... (t) re-div-ivu-s, di-u, inter-di-u-s, nu-di-u-s, di-es, di-ur-nu-s, ho-di-er-nu-s, di-es-piter, nun-di- na-e. (e = al) Sk. dev-a-s, a god. (ei) 6£7of, E. L. deivos, deiv-a. (i) div-us, di-us (weak- ened form de-us), Di-a-na, I-anus for Di- anus. In bi-du-um, tri-du-um, &c., du is a weakened form of div. (1) i-t-er, ad-i-tu-s, &c., in-i-t-iu-m. ... (e = ai) Sk. e-7ni. (oi) ol-juL-o-g. (ei) el-fx-ij E. L. ei-re, (i) i-re, I-bo, i-vi. ... (1) i-s, i-b-i, i-dem... . (ei) E. L. ei-eis. (i) T-dem, (1) qui-es, qui-e-sc-ere, qm-e-tus .... (ei) Keijjiaif E. L. cei-vi-s. (i) ci-vi-s. (1) li-n-ere (-ire), li-tu-s, li-tu-ra. (ei) E. L. lei- t-er-a. (i) li-mu-s ... li-m-ax, li-n-ea, li-t- er-a or li-tt-er-a. (G. schleiin^ Engl. sli7Jie.) (i) -spTc-ere -spec-ere ... (ei) E. L. peic-u-s. (i) pic-u-s, su-spic-io ? (G. specht. Engl, wood- pecker ; pie^ (1) tri-bu-s, tri-dens ... ter, tre-centi. (ei) rpelQ. (i) tri-s, tri-ni, trI-du-um. (i) fid-es, fid-eli-s, per-fid-u-s (oi) E. L. Jbid-u-s, foedus, foed-er-a-tu-s, iriiTOLda. (ei) TTSLdu), E. L./eidus. (i) fid-ere, fld-us .... {%) hb-et. (oi) E. L. loeb-er. (ei) E. L. leib-er. (i) llb-er .... 6 Latin Soundlore. c) Primitive or Italian a, sometimes represented by e, 6, or u, is strengthened into a, which sometimes sinks to e or 6. (i) a (e, 1, 6, u). (2) a. (3) e. (4) 6. ak, sharpen ag, drive . bha \throw bha-s flight bhrag, break kar, wa^zt , gan, t^<^^/ . gna, know kal. lab, slip ma, measttre ^diVy fastejt rag, direct . sa, j'^?'Z£/ sna, bathe l) ctfc- ac-u-s, ac-u-ere. (2) ac-er. (4) aKOJKrj, wk-vq, oc-i-or I ) dy- ag-ere . . . ag-i-li-s .... (2) amb-ag-es, ac-tu-s. (3) ^^-h orTpaTTiyog .... (4) paedagog-u-s. i) fe-n-es-tra, fa-t-eri, fa-t-uu-s, infi-ti-a-s, fax fac-, fac-ere ... f ac-ie-s, fa-ber, 0d-oc, 0d-mt. (i) fa-s, fa-s-ti, fa-ri, fa-bula, fa-ma, fa-c- und-us ... fa-tu-m, fa-nu-m. (3) fe-t-i-aU-s, /e-s-ia-e = {e-r-ia.-e, fe-s-tu-s, ^J//Lte, f^firj, i) fra(n)g-ere ... frag-ili-s, frag-or, nau-frag-us, (2) refrag-ari, suf-frag-iu-m ...fr ac-tu-s. (3) {YQg-'i, Pprjy- .... (4) eppioya. i) car-ere. (2) car-us. i) gen-us ... gen-ui, yep- yov- gen-er, pro-gen- ie-s, gen-iu-s, in-gen-iu-m, indi-gen-a, gen-i- tor, yvvi]. (2) gna-tu-s, na-tu-s ... na-t-io, natura, gna-vu-s, i-gna-vu-s .... (3) yvijaiog, i) no-t-a, no-t-are, c6-gm-t-us .... (2) gna-ru-s, i-gna-ru-s, na-r-r-are. (4) no-tu-s, i-gno-tu-s, igno-r-are ... no-r-ma ... no-bih-s ... no- men, a-gno-men, i-gno-min-ia ... eynoKa. 1} cle-p-ere (cli-p-eu-s), cel-la, oc-cul-ere, cu-cul- lu-s, -cil-iu-m, c-la-m, kKotttj, (2) cal-ig-o. (3) eel-are. (4) KXojxp, i) lab-are, lab-e-facere. (2) lab-i lab-es. i) tna-nu-s, juerjoeli^, me-tru-m, me-d-imnu-s, me- t-ere, mo-diu-s, mo-d-u-s, mo-d-er-ari, mod- es-tu-s (2) ma-ne, im-ma-ni-s, ma-tu- ru-s, Ma-tu-ta. (3) me-ta, me-t-iri ... me-n- sa, me-n-sura, me-n-si-s, se-me-s-tri-b .... (4) mo-s mor-, mor-osu-s. i) pac-i-sc-i. pa(n)g-ere, pe-pig-i, Trdy-. (2) pax pac-, pac-are, re-pag-ulu-m, pag-u-s, pag-ina, com-pag-es, pro-pag-o. (3) -pegi, 7r?;y-. i) reg-ere, reg-io, -ngere, rog-u-s. (2). Sk. rdjan. (3) rex reg- ... reg-ula (From regere perhaps rig-ere, rig-i-du-s ... .) i) sa-tu-s, sa-t-io (se-rere) .... (2) Sa-t-ur-nu-s. (3) se-vi, se-men .... f) na-t-are. (2) na-re ... na-r-i-s, na-s-u-s, nSs- turt-iu-m (nasum torquens), (3) rfjcroc. § 12. Vowelstrcngthening in Root-syllables ^ etc. sta, stand . 17 star, strew . (i) sta-tor, sta-ti-m (but E. L. std-ti-m), sta-ti-o, sta-tu-s, sta-tu-ere, sta-tu-a, sta-bulu-m, sta- bili-s : ste-t-i, super-sti-t- (2) sta-turu-s, sta-tura, sta-men, and in Conjug. sta-re, sta-bam .... (3) ori/vai .... (i) ster-(n)ere (f^rop-), (2) stra-vi, stra-tu-s, stra- men ... stla-ta, stla-t-ariu-s, (4) GTpio-wi-vat^ arpw-fxa* d) The following are strengthened from (i) a (e or 1) to (3) e (I). (i) (h)er-us, (h)er-a, (h)er-c-i-sc-ere, hir-und-o, hir- ud-o, E. L. Mr (x^p-) (3) (h)er-es, (h)er-ed- (i) Cer-es, cre-are, cre-sc-ere (3) cre-vi, in-cre- mentu-m .... (i) sed-ere ... as-sid-uu-s, (3) sed-es, sed-i, sed- ulu-s, sid-ere... . (i) ser-ere ... ser-ie-s, ser-a, ser-tu-m, (3) ser-ia, Pr. har, seize kar, create sad, sit S3x^join skar, sever stag, cover var I cover val f choose ser-u-s, ser-iu-s. (i) cer-(;^)ere ... cer-tu-s, ... scre-are, (3) cre-vi, dis-cre-tu-s ex-cre-mentu-m, cri-bru-m, cri-men, dis-crl-men .... Kpivu)' (i) (TTEyiOj Tey-ocj teg-ere, teg-es, (3) teg-ula, tex-i. (i) val-lu-m, ver-eri, vel-le, val-ere, val-idu-s, (3)' vel-u-m, vel-are ... : ver-us, ver-ax ... • From another root val, implying motion, come (i) vol-are, (3)vel-ox, vel-es, vel-it-ari. e) The following are strengthened from (i) Pr. a (6) to (4) 0. Pr. sar, de whole svan, sound svap, sleep svar, sun vak, call (i) sal-us sal-u-t-, sal-u-ber, sal-vu-s... , (4) sol-ari, sol-a-c-iu-m. (i) son-u-s, son-are ... (4) per-son-a. (i) s6p-or ... som-nu-s, (4) sop-ire. (i) ser-enu-s, (reX-a^, aeX-rivrj, ^eip-Log, (4) sol (Sk. siir-a-s^. (i) voc-are, (4) vox voc-, v6c-ali-s (con-vlc-iu-m ?). /) C. cites also many examples of 1 (e) rising to i and of u rising to u in Latin without diphthongal accretion. Such are (a) liqu-ere ... liqu-or, (^) liqu-i, liqu-or. (a) sec-are ... (/6) sic-a, sic-ar-iu-s. (a) stil-u-s, stim-ulu-s ... instig-are. («) i/y-po-c, ifi) u-v-ere, u-m-or ... * The examples in pp. 14-17 are selected from a large number in Corssen's great work, I- 348-550- The instances cited are the most important of those in which the short as well as the long vowel occurs in words of classical use. Forms from old Italian dialects and from other languages are here given only so far as they illustrate diphthongal strengthening. Other roots of great interest will be found in Corssen's pages ; as Par. ■pnYyJill'. whence plere, plus, plenus, populus, plebs. Mar, glitter ; whence mare, marmor, Mars, Mamers. Mar,ya^/,?: whence marcere, mors, mori. Bhu, be horn : whence fu-, fore, fe-, fetus, femina, fecundus, fenus. C 1 8 Latin Soundlore, * § 12.« B) Vowelstrengthening in Suffixes, Case-endings, and Personal- endings, will appear in the Sections which treat of Declension, Con- jugation, and Derivation : hon-os honor-is, Cer-es, pulv-Is, nub-es, lig-on-is, matri-mon- ium, matr-on-a ; — mens-ar-um, de-or-um ; — am-as, am-a- mus, am-a-ris, fu-e-runt, fu-i. Disyi- xiv. Vowelstrengthening in Perfects. Per- i) Most Verbs with vowel character a, 1, o, and some with e, fects. formed the Perf. in vi, and lengthened the character ; na-vi, ne-vi, no-vi, li-vi. 2) Of u-verbs, C. says that their Present-stem anciently received the strengthenings ou, u, before it was weakened into ii ; and that the Perf passed through the forms -uvi, ul before it was weakened into ul. Thus plouo, pluo became pluo : and pluvi, plui be- came plui. The only exceptions are batuo, -gruo, metuo, ruo, which seem never to have lengthened u before a vowel : and fuo, which in the Perf. became foul (poet, fovi), fui, and ultimately fui. 3) As to the formations caveo cavi faveo favi paveo pavi foveo fovi moveo movi voveo vovi C. thinks that (to avoid the concurrence -vui) ii was thrown out, and the root-vowel then strengthened : cavui, cav-i, cavi. 4) As to the following three — sedeo sedi ; video vidi (E.L. veidi) ; venio veni he assumes the existence of old forms sed-ere, vid-ere, ven-ere, which in the Perfect were strengthened as the following Consonant Verbs : fra(«)g- fregi leg- legi (lexi) ed- edi fod- fodi fu(;/)d- fudi cap- cepi rvi{m)^- rupi scab- scabi em- emi fac- feci iac- ieci li(;/)qu- iTqui \i(n)c- vici ag- egi He brings reasons against the common assumption that in such Perfects the long vowel compensates for a lost reduplication. And, in fact, the practice of vowelstrengthening in Italian Soundlore is so well established, that no such assumption is necessary to explain the quantity. Yet fefici (O.fefdci) is known as an older form than feci, while pegi from pango and tudi from tundo appear as secondary forms for pepigi and tutudi. We may also compare Greek forms, as ayriyov or i^yayop with egi, IPiw^rj with edi, oida with VI di or veidei, and be led to doubt whether, in some instances at least, a loss of reduplication may not have caused the root-vowel to be strengthened in Latin. Com- XV. Compensation. Compensation is usually said to happen when a naturally short vowel is lengthened in order to maintain the quantity of a syllable after the loss of a consonant: vid-sum, visum. But such com- pensation is not always made: seget-s, sages. i Nasalisaiion, Vowclwcakening. xvi. Strengthening of the Present Stem in Verbs by Insertion of a Nasal. Some Verb-stems have the short vowel of their Present-stem strengthened by Nasalisa tion ; that is, by adding n to the Stem- vowel before a Guttural or Dental, m before a Labial character : Nasali- sation* frag- lug- ng- hqu- mig- nigu- pag- frango iungo ling- linquo mingo ninguo pango pig- pingo pug- pungo strig- stringo tag- tango VIC- vinco fid- findo fud- fundo scid- lud- pis- cub- lab- rup- scindo tundo pinso cumbo lambo rumpo As the Nasal for the most part disappears in Derivation, it was evidently not so strongly sounded as in modern utterance. Guttural n is called by grammarians nadulterinum. (On the strengthening of the Present-Stem by Suffixes, see § 52.) xvii. Vowelweakening. ♦ 1. Pr. roots are formed with each of the Pr. vowels, a, i, u : but those with a are by far the most numerous. 2. The standard vowel a is weakened into i and u in all Aryan languages : but in those which have e and 6 the weakening of a into 1 passes through e, and the weakening of a into u passes through 6 : Pr. sama E.L. semo-l C.L. simu-1 3. Italian dialects shew such weakening largely ; in Root-syllables, in Suffixes, and in Endings of Case and Person. 4. The general object of all such changes is Euphony (fv^uu^/a), the more easy and convenient utterance of the sounds of speech. 5. In pursuit of this object certain principles are applied ; among them Assimilation and Dissimilation, hereafter noticed ; also Selection, which occurs when a certain vowel is chosen as the most suitable before a particular consonant. Thus, v has a prefer- ence for o ; 1 and the labials chiefly for u ; r for e ; n and t for i. Grouped consonants often prefer e : liivis volt epopa peptri becomes lovis vult upupa peperi Menerva becomes Minerva aletem — alitem voile — velle faciundus — faciendus Note, E is the easiest and smoothest Latin vowel, being neither so sharp and thin as i, nor so flat and thick as 6 and ii. Hence it prevails as a final vowel, and in several instances is so used when final consonants are cast off: venere for venerunt ; utare forutaris; dictatore for dictatore-d or dictatori-d. Also for 1 final in Neuter Nouns : mare for mari-. But when Masculine or Feminine Nouns drop final n, the stronger vowel o becomes final in Nom. Sing. : homo (homon-, homin-), virgo (virgon-, virgin-). c 2 Vowel Weak, ening. 20 Latin Sotmdlore, § t2. Weak- xviii. The Vowel a and its Weakenings. enings of a. A) In Root and Stem syllables. 1) Corssen (II. 6) cites about 270 Latin words which have kept Pr. a in the syllable of the root or stem : acus, aqua, traho, daps, labor, pateo, madeo, maneo, ango, pando, amo, sal, valeo, palleo, malus, careo, carmen, hasta, aveo, gravis. 2) He cites about 215 words which have weakened Pr. a to e in the syllable of the root or stem : dScem, neco, equos, sequor, tego, tepeo, febris, peto, edo, senex, fremo, mel, queror, tero, vespa, severus. And others which have passed from a through e to i : digitus, ignis, quinque, pinguis. 3) He cites about 190 words which have weakened Pr. a into i o in the syllable of the root or stem : voco, mox, loquor, rogo, opus, 6b, nota, fodio, tono, vomo, mola, orior, voro, post, novem. And others which have passed from a through o to u : nummus, unguis, fungus, multus, culmen, vulnus, a) Pr. & is weakened to e and 6 in some roots ; nex, neco, noceo ; tego, toga, bene, bonus ; mens, moneo ; pendo, pondus. fero, fors, fortis ; verto, vorto ; volo, velle, volt (vult). ' preces..., procus, posco; cello, coUis, columen. V) Pr. a, kept in Latin, also becomes S in fatisco, fessus ; gradior, gressus. c) Pr. a, kept in Latin, also becomes 6 in apiscor, opus ; scabo, scobs ; pars, portio ; faveo, foveo. d) Pr. long a becomes 5 in some words and many suffixes : donum, vox, mos : -top, -ds, -OP, -mon, -an, and Imperative -to. In Suffixes. It may be stated as a general rule that Latin suffixes with the vowels e, 1, 6, u, are weakened from Pr. suffixes with a. Exceptions are very few : tl- in such words as menti- parti- potJ-. nl- in such words as igni- pani-. VI- in words like ovt- avi-. tii- (sii-) in Supines and Nouns, as statu-, dictu-, casu-. AH which are in Pr. form. C) In Cases and Personal Endings. Cases (except the Locative Singular) and Personal Endings, with vowels e 1, o, u, are for the most part weakened from Pr. forms with a. ' See §§ 20, 39, and Schleicher, Vergleich. Gramm. def Indogerm. Spr., § 205. i g 12. Vowehveakening with Selection, 21 xix. Weakening into 6 as influenced by Se- ^^^^^^ lection. 1. V following has determined Pr. a to become 6 in novem, novus, 1 6 vis, ovis. 2. V preceding has probably done this in voco : while in vol volvo, volnus (vulnus) andvomo, the consonants which follow may also have had influence. E. L. has voco for vaco, voto for veto, vorto, voltur. M. Lucr. i. 20. 3. Pr. sva is changed to so in soror (for svasar, ^sister'), sopor, socer, sonus, sol, sodalis; and has passed to su in su-sur-rus. See p. 17. C. II. 64. 4. It shews a preference for o before it in many words : dolus, solum, solium, tollo, &c. but especially in E. L. suffixes : poC'dl'OMj tab-ol-a, Pseud-ol-us^ po-pol-us, which afterwards changed 6 into u. 5. Inner r often prefers otoii: fore, forem from fii-o; so ancora from Gr. ayKvpa (but generally Greek v was kept before p: pur-pur-a): especially in the Suffix of Neut. Substantives with Nom. S. us or ur : corp-us corp-6r-, eb-ur eb-6r-. Some keep ur- : fulg-ur-, gutt-ur-, murm-ur-, Sulf- ur- ; and the Masc. words aug-ur-, turt-ur-, vult-ur-. (But most Neuters in us are inflected by er-. See p. 25.) 6. The Comparative Suffix was anciently ds or-, then or or-, for all genders : finally, and in C. L., it became M.F. meli-or meli-or-, N. meli-us meli-or-. Note p. 42. XX. Weakening into u as influenced by Se- Seiec- lection. 1. C. says :^In Latin root-syllables, suffixes, and flexional end- ings, XL has arisen generally from 6.' a) before s and m final : deus {Qe6q\ genus (ytvov), bellum (bellom), fihum (filiom). d) before inner 1, or a labial : populus, upupa,'columen, Hecuba. c) before grouped consonants, the first of which is a Liquid, Nasal or Sibilant : pulsus, palumbes, fungus, rursum, luscus. 2. About 230 B.C. the 6 of case-endings generally passed into : hut 6 was kept in some instances : a) hoc, tot, quod, quot, always. So com- con-. 22 Latin Sotmdlore. - ^ 542, /3) after u, v, as late as the Augustan age : equos, equom, servos, servom, aevom. So quom. The Emperor Claudius seems to have promoted the use of the combinations uu, vu, which in Republican times were generally avoided. See C. 11. 97-101. 3. Rustic dialects kept o frequently : hence it returned into use in L. L., and reappears in modern Italian : popdlo^ secolo^ 4. Selection of u appears A) before Labials and 1 : 1) in place of o : hiimanus utrubi consul umerus bubile adulescens nummus bubulcus epistula voliimus bubus exsul quaesumus riibigo titulus sumus iipilio singuli Bovile is another form for bubile. 2) in place of Gr. a, c : humus (xct/^af) scopulus (aKoirtkoq) pessulus (Trao-o-aXoc) Siculus (Si/ce\oc) 3) in place of a, e : a : occupo, aucupor, contubernium ; e : quincuplex. 4) as middle sound, approaching to 1 : clupeus or clipeus : lacruma or lacrima. See xxi. Note I. When 1 or e came before the suffix 616- 61-, 6 was not changed into ii : (1) fili-61u-s, basi-61u-m, Cori-ol-i, vi-61-entus. (2) alve-olu-s, lacte-61u-s, laure-ola, Pute-61-i.^ Note 2. When v came before 61, 6 was kept as late as the Au- gustan age, after which it often became u : parv-61u-s, parv-ulu-s : serv-61u-s, serv-ulu-s. But friv-61u-s was never changed. * As to the dialectic use of O and u in final syllables, C. says : 1) Lat. and F. received O as the pure Gr. o. O. and S. as O inclining to U. ' ' U. as a middle tone, or 11 inclining to O. 2) Lat. changed O to U about 300 . . . 200 B.C. F. still earlier. O. about 300 B.C., but before m not till 130 B.C. S. before 200 B.C. V. never. New U. returned from U to O between 300 and 130 B.C. MSS. shew formid-ul-osus and formid-ol-osus, sanguin-ul-entus and sanguin-ol-entus, vin-ul-entus and vin-ol-entus ; the forms with ul- having the advantage. Somn-ul-entus ii decidedly better than somn-ol-entus. ^12. Vowelweakcning with Selection. 23 , B) Before grouped consonants, of which the first is a Liquid, Nasal, or Sibilant : I. In place of o. 1) pullus, cucullus, and the Deminutives ampulla, homullus, Catullus, Marullus, &c. But before 11, e is more frequent than u. See xxi. D. 6. fulcio ... , hiulcus, pulcher sulcus Vulcanus ; fulgeo mulgeo..., vulgus ... . culpa . . . , bulbus. adultus ... , cultus multa pul(t)s, ultra ... vultur, vultus, vult : — vulsi. culmen, fulmen, ulmus ; ulna. Fulvius, pulvis, vulva, ulva. Here too ol after v held its ground long : Volcanus, volgus, voltur, voltus, volt, volsus. 2) amurca, furca, urceus, murmur, furnus, eburnus, purpura, Surrentum, ursus, rursus. 3) umbo, nummus, aerumna, alumnus, autumnus, columna, Clitumnus, Vertumnus. 4) uncia, uncus, hunc, Aurunci ; fungus, unguis .... The Demin. suffix -unculo- : ranunculus, virguncula. 5) Promunturium ; nuntio (noventio). Corss. I. 5 1 ; nundinae. The Personal Ending -unt was anciently onti- ont, as ecfociont for effugiunt on the Columna Rostrata, con- sentient^ dederont^ &c., on old inscriptions. The classical form of 3d Pers. PI. Perfect -runt for -ront first appears in the Senatusconsult. de Bacc. B.C. 186, consoluerunt; the weakened form in -re somewhat earlier, fee ere. When u or V came before -ont, o was kept to a later time. Thus in the MSS. of Plautus appear ruont, per- pluont, vivont; and in Lucretius loquontur, disso- luont, vivont. The inner suffix -unt- {ovt-) appears in a few words : so euntem ... , chironomunta (Juv.) ; Acherunta (Plaut. Lucr.). The suffix -un-do- (for -on-do-) appears in bar-undo, hir-undo, sec-undus, ori-undus, rot-undus, fa-c- undus, fe-c-undus, vere-c-undus, furi-b-undus, tremi-b- undus, vaga-b-undus ; and in Gerundive Participles.* * Of the Gerundive forms -ondu-s, -undu-s, -end-us, C. (I. 180) shews that (1) No existing E. L. Inscrr. contain -ond-us ; but, as it was the tendency of L. L. to resume the o of E. L., and in L. L. appear such forms as secondus, verecondus, while Italian also has secondo, rotondoy it may justly be assumed that -on-do- was the first weakening of Pr. -an-d-ya. (2) The forms -undus -endus appear side by side in E. L. and R. L. to the Christian era. So in Plant, Ter. Lucr. ; in the Senatuscons. de Bacc, the Lex lulia. (3) The form -endus prevails in prose : but Sallust likes -undus. Cicero, Caesar, Livy, use it often, chiefly in io-verbs of the 3rd as well as 4th Conj. : moriundum, partiundus, &c. It prevails especially in legal and statistic phrases : rerum repetundarum, iurc dicundo, belli gerundi, agris dividundis. 24 Latin Soundlore. § 12. Rarer forms are frundes for frondes, frunte for fronte, dupundius for dupondius. Later language resumed o : ^ It. fronde, fronte, pondo, mondo. 6) The Demin. forms arbuscula, corpusculum, rumus- culus. aplustre, indu-stria (endo) : also arbustum, onustus, robustus, venustus; which C. would derive from weakened forms arbus, onus, robus, venus. 2. The words in which u appears to represent e are few : as malsum (jjiiXi) ; sepultus from sepel-ire ; urgeo (tjoyo)) .... Note, Long u for Pr. a or its substitute 5 appears in various suffixes : -turo-, -tura, -suro-, -sura, -uno-, -una^ -uco-, -uca, -uceo-, -ucio- : i-turus, prae-tura, pas-surus, men-sura, Nept-unus, fort-una, ^ cad-ucus, fest-uca, pann-uceus, Vin-ucius. compared with prae-tor (anc), patr-onus, fer-ax, fer-ox, mer-acus, clo-aca, gallin-aceus. xxi. Change into e as influenced by Selec- tion. It has been shewn that A) e has affinity with r, B) e is a convenient vowel for the close of words and for finai syllables. Also it must be noted that C) e has affinity with the combinations st, ss, U. Z>) e is a convenient letter for the syllable of Reduplication. These causes determine a great number of instances in which e appears for other vowels in suffixes and endings. A) e chosen with r. i) In Decl. I and 2, before the suffix po- pa-, e takes the place of Pr. a or of o, sometimes of u : Examples are- Words which retain vowel and suffix throughout : numeru-s, umeru-s, uteru-s ; camera, littera, tessera ; and the Fern. Adjectives, lacera, libera, misera, tenera, altera. • Words which drop the vowel of the suffix in Nom. and Voc. Sing. Masc. : gener, puer, socer (Ifcvpoc), vesper, lacer, liber, miser, tener, alter. 12. Vowelweakening with Selection. 25 Words which drop the vowel of the suffix in Nom. and VoCo Sing. Masc, and drop e in all their other forms : ager, aper, faber, magister ; Afer^ ruber, niger, uter. In some of these (as magister and uter) e represents Pr. a, while in others it is perhaps a transposed repre- sentant of the dropt 6. In some words with suffix ro- ra- Greek a was kept befoire r : , cantharus, cithara. 2) Similarly, in Decl. 3, e comes before r in many forms : Words in er er-, which keep e throughout : career, cadaver, piper, papaver, tuber : celer, degener, pauper, uber. Words in ter, cer, ber ; which keep e ( = Pr. a) only in Nonu and Voc. Sing. Masc, dropping it in the other cases : frater, mater, pater, venter, imber ; acer, alacer, saluber, puter, Silvester. Numerous words of Decl. 3, which take Nom. S. iis ( = ds), have in the Oblique Cases the suffix er- : fun-er-is, gen-er-is, op-er-is, Ven-er-is, vet-er-is. but some keep 6-r- : corp-6r-is, frig-6r-is, pect-6r-is, a few use er- and or- : pign-iis, pign-er-is and pign-6r-is : fenus, fen-er-is and fen-6r-is ; temperi for tempdri : whence tempero, temperies. Vomer or vomis, Gen. vom-er-is, points to an original form in es, which sometimes weakens the consonant and be- comes er, sometimes weakens the vowel and becomes Is : see xxii. 2. and compare the forms Ceres Cerer-, cinis ciner-, pulvis pulver-, 3) Verbs having er in their root do not weaken e into 1 in compounds : affero, congero, desero, puerpera. for the same reason peperi not (pepiri) ; repperi not (reppiri). similarly the compounds of iuro become de-iero, pe-iero, weakening u into e. When the Perfect receives suffixes which begin with p, i is changed into e or e : (E. L. dedi-sont dede-ront) dederuntordedere. (E. L, dedt-so)y dede-ro. 26 ' Latin Soundlore. . - § 12. B) B has a tendency to take the place of other vowels in final syllables before weak consonants, n, m, s, t. 1) In Decl. 3, the Norn. S. -en (for Pr. -an) becomes -in- in the Oblique Cases : pecten, agmen, crimen, oscen, tubicen. Gen. pectin-is, agmin-is, crimin-is, oscin-is, tubicln-is. 2) a. Septem (Sk. saptan, Gr. tVm) ; novem (Sk. navan^ Gr. tvvea), decem (Sk. das' an, Gr. Uku). /3. In Decl. 3 -em is the Accus. S. Ending of Cons. Nouns : reg-em, virgin-em, passer-em. it prevails in I-nouns against -Im: civ-em, font-em, serpent-em, but -Im is retained by some : bur-im, sit-im, tuss-im, vim, while others use both forms. See § 24, 5. febr-im febr-em, messim messem. Tim is kept by the numerous Adverbs having that form : as ad fatim, partim. But saltem or saltim. M 2LS 3, final letter faded in L.L. as it had faded in U., and passed out of use in modern Italian, leaving e final generally : se^/e, nove, &c.; hut unded, doded, ... from undecim, duodecim, ... 3) On such forms as nub-es, sed-es, see § 20, 24. Some I-nouns have two forms of Nom. S., -is and es : fel-is fel-es, vall-is vall-es, verr-is verres; but s, like m, disappeared in L.L. and Italian, leaving e final : nude, valle, &c. 4) In old Italian dialects, except O., also in E. L., final t in Verbal forms was weak and sometimes disappeared. Before it the Perfect character i was sometimes changed into e : at a later time to the middle sound ei : finally in classical times settling into I. Thus are found the various forms : (dede, dedit, dedet, dedeit), dedit. In L. L. and in Italian, this t, like m and s, disappeared again, leaving final e ; disse^fece, C) S has a tendency to become itself a final letter in the place of other vowels. 1) In the Voc. S. of 0-nouns it supersedes 5 : domine, lupe, Romule. 2) In the Neut. S. Nom. Accus. of I-nouns it supersedes 1 : mare, rete ; triste, necesse. 3) When final consonants are cast off: q\x\ViC{Mc{Sk. panc'Uy Gr. wivri). §12. Vowelweakening with Selection. 2/ ille, iste, ipse (illus, istus, ipsus). Abl. S. of Decl. 3: quaestore (quaestorid or quaestored). -re for runt in Perf. dedere (dederunt). -rS for -Pis in 2nd Pers. S. Pass. : loquare for loquaris. -ve for vis in neve, sive. mage for magi s; pote for potis. In L. L. instances occur even of a Gen. S. in e for is. By this gradual rejection of final consonants the classical system of case-inflexion was broken down and the uniform declension in- troduced which prevails in modern Italian. D) s has a tendency to take the place of other vowels before grouped and double consonants. 1) JB appears before x ( = cs, grs) in the final syllable of Nouns of Decl. 3 which are inflected with the suffixes ic- ig- ( = ICO- iffo-, as explained by Corssen) : codex, cortex, imbrex, remex ; simplex, supplex. Gen. codicis, corticis, remigis ; simplicis, suppllcis. 2) S appears before ps, bs in the final syllable of Nouns of Decl. 3, which are inflected with the suffixes ip- (up-) ib- : such are municeps, auceps, caelebs, particeps. Gen. municipis, aucupis, caelibis, participis. Compounds of caput, with Nom. -ceps for -cipit-s, have Gen. -cipitis ; praeceps, Gen. praecTpTtis. 3) When a Noun with that suffix ti- (which appears in hos-ti-s, tes-ti-s) would have the accent on an ante- penult syllable (ala-ti-, equ6-ti-), the vowel of the penult is weakened usually into 1 (aliti-, equiti-), sometimes into e (abieti- segeti-). The i of the suffix being dropt, the forms then become (alit- equit- seget- abiet-) : and when the Nom. S. is formed by the addition of -s, they become (alit-s equit-s seget-s abiet-s) : but, e being preferred to 1 in a final suffix, (alit-s equit-s) become (alet-s equet-s). After which, by the rule of euphony, the dental falls out before s, and the Nominatives then become ales, eques ; Gen. alit-is, equit-is; seges, Gen. seget-is; teges, Gen. teget-is. but abies, aries, paries. Gen. abietis, arietis, pa- ri etis, on account of i preceding. Note, In this class, the vowels e, 1 generally represent Pr. or Latin a (see above), but in a few % is the root- vowel : comes, comit- (root 1, to go). 28 . Latin Soimdlore, ; - § 12. In pedes pfedtt-, probably also in seges, teges, the vowel is adopted by analogy, forming a suffix i-t- or e-t-. See Footnote, p. 30* 4) The same principle applies to a few words derived from sed- sid-, to sit (Pr. sad) : (obsed-s) obses obsid- ; (praesed-s) praeses, praesJd-; (desed-s) deses desid- ; (resed-s) reses resid-. 5) S before nt appears in the suffix mento- : ar-mentu-m, la-mentu-m, monu-mentu-m. and for Greek a in talentum, Agrigentum, Tarentum. 6) B is frequent before 11 : cello, pello, vello, velle, and the Demin. forms : puella: but in these i is also used : sigillum. In other groups with 1 the vowel u prevails, see xix. : but e is not excluded : celsus, excelsus. 7) Equester, pedester, for (equet-ter, pedet-tef). 8) (fat-) fessus ; (grad-) gressus. 9) The Neuter suffix (os) us weakens its vowel into e before another suffix beginning with t : fun-US fun-es-tus ; scel-us, scel-es-tuSc temp-US temp-es-tas ; intemp-es-tus. The existence of an old Neuter Noun modus is shown by mod-es-tus ; mod-er-ari : so the Masc. Noun honos forms hon-es-tus, hon-es-tas : but o becomes u in ang-us-tus, aug-us-tus, on-us-tus, rob-us-tus, ven-ua-tus. maius, mai-es-tas, is like temp-us, temp-es-tas. C. forms pot(i)os, pot-es-tas : others poten(t)s (potent- tas) pot-es-tas % he cites Prasrn-este as Superl. from a supposed (prae- no-), meaning * the town on the highest prominence^ 10) The comparative forms mag-is-ter, min-is-ter, sin-is- ter, in L. L. appear with es for Is ; whence \X.2X\2,Ti maestro. In some other words also, as an teste s, L. L. writes est- for 1st-. Modern Italian is not uniform in the choice between e and l. We find fermo^ selva^ segno ^ irenti; but principe, sinistra^ vittoria^ carissimo: E) On the use of e for fi, 6, u in the reduplicated syllable of Per- fects see xxv. g 12. Vowelweakening with Selection. 29 xxii. The Selection of i. aonofi. A) The thinnest and sharpest vowel i has a strong affinity with dental consonants ; chiefly with n and s, but also with t and d, E) Hence it is largely used as a vincular vowel, linking stem with suffix and suffix with suffix. C) The existence of a middle sound between 1 and u caused the orthography of many words to fluctuate. A) I. Affinity of i with n. I represents Greek a before n in balineum, bucina, fascino, machtna, pattna, trutina, ^ It represents Greek i before n in adamantinus, coccinus, coccineus, crystallinus. It stands before the suffix no- in numerous Latin words : * fisclna, fuscina, pagina, sarcina, pampinus, sucmum, fagtnus, fagineus, geminus, myrrhinus. In mino- (Pr. mana) and tino- (Pr. tana) : terminus, femina ; fruimino, amaminor. crastinus, diutinus, pristmus. In the suffix in- (Pr. an L. en, en) before vowels : pect-in-is, sangu-in-is, osc-in-is. In the suffix in- (Pr. an L. on, on) before vowels : ^ hom-in-is, marg-in-is, ord-m-is, virg-in-is, ApoU-in-is. In the suffix min- (Pr. man L. men) before vowels : flu- mm-is, no-mm-is, nu-mln-is. A striking instance of the affinity of 1 with n appears in the fact that it was inserted in the Greek word \i'va^ which so became mina. Similar insertions occur in Daph-i-ne^ luc-i-nus or lych-i-nus (M. Lucr. p. 211), gum-i- nasium probably in Catullus. So the affinity of u with m is shewn in the occasional forms drac-u-ma for lpa)^\ir]^ Alc-u-mena, Tec-u-messa, &c., and with 1 in Aesc-u-lapius, Herc-u-les. Minerva, anciently Menerva. Though e prevails before grouped consonants, yet there are many instances of it being sharpened into i before n with another consonant : intus, inter, indu- . . . quinque . . . tingo . . . vindico . . . so when n follows another consonant : ignis, pignus, signum, tignum. 2. Affinity of i with s is shewn In the forms cinis (ciner), cucumis (cucumer), pulvTs _ (pul-ver), vomis (vomer), pubis (puber) : also aci- pensis (acipenser). See C. II. 278. * In fact the suffix no- takes, in true Latin words, no short vowel but i before it. Such words as balanus, cottana, platanus, raphanus, Rhodanus are not native of Italy. 30 Latm Soundlore, ] . §12^ In the Gen. ending -is (Pr. as). In the occasional use of l-sc- for e-sc- in Inceptive Verbs : lucisco for lucesco, 3. Affinity of 1 with t is shewn In the adoption of i before many Verb and Noun suffixes beginning with t : ag-i-to, ag-i-te, ag-i-tis, gen-i-tus, gem-i-tus, domi-tum, meri- tum, veti-turus, doli-turus, fru-i-turus, gen-i-tor (but gene- trix), habi-tare, strepi-tare : — laeti-tia, plani-ties, veri-tas, alti-tudo, peni-tus, largi-ter, sempi-ternus. 4. Affinity of 1 with d is shewn In the adoption of 1 before the suffix do- : candi-dus, torp-i-dus, flu-I-dus, viv-I-dus. herbi-dus, gravi-dus, morbi-dus, gell-dus. Note. When an E-verb forms a Substantive with suffix d-on- ^'Xn-, the vowel before that suffix is e : albe-do, dulce-do, grave-do : but lib 1- do, by assimilation. B) Use of i as a linking Vowel.^ I. The large use of i before suffixes beginning with n, t, d, and its own aptitude for this purpose, led to its adoption before many other suffixes as a link-vowel in the place of others : as before co-, c-un-do-, culo-, cro- ; b-un-do-, bulo-, bill-, men, men-to-, monia. (Verbalia) ali-ca, velli-co, medi-cus, rubi-cundus, cubi-culum, ridi-culus, veh-i-culum, pudi-bundus, fur-i-bundus, pati-bulum, cred-i-bilis, terri-bilis, flexi-bilis, spec-I-men, al-I-mentum, quer- i-monia. (Denominativa) auli-cus, belli-cus, annt-culus, ludi-cer, currT- culus, ani-cula, aegri-monia, caeri-monia. But Verbal a is kept : ira-cundus, caena-culum, vaga-bundus, ama-bilis, grava-men, sacra-mentum. Sometimes e : vere-cunduSj fle-bilis. * Corssen is right in principle, when he considers this T to be a weakening of the final Towel of Stems with vowel-character ; as in aulT-cus from aula; bellT-cus from belle- ; ridi-culus from ride- ; am-cula from anu-. But he seems to go back too far when (II. 314 and elsewhere) he speaks, for instance, of the 1 in regimen as weakened *from the original final a of the 3rd Conjugation.' He might surely have applied here and in other Derivatives of Consonant Nouns as well as Verbs the principle which he admits, for ex- ample, in ped-es, ped-it- from the root ped- (Pr. pad, Gr. ttoS-), and in the use of the suffix i-co- (II. 211. 205) ; namely, that the usage of vowel-stems, which adopt! so gene- rally as a light link-vowel, has thus created a uniform sujffix (einheitliches Suffix) applied, by linguistic analogy (Sprachbewusstsein), to Consonant stems also. This is, in fact, all that is meant when the use of vowels (1, u, e) is cited in this Grammar as * vincular : » and in this sense the term will be still kept as convenient. The same convenience recommends the term * Clipt Stem ' to express a vowel-stem without its vowel character. But * mord ' is in fact the root of mord-ere. Hence, to say that momord-i, morsum, come from a theoretic verb mord-ere, as C. does, and to say that they are formed from the Root of the extant Verb, are but two ways of saying one and the same thing ; and the latter is the shorter way» § 12. Vowelweakening with Selection. 31 2. A similar adoption of x is frequent in Compound Words at the close of the prior element. (i) terri-gena, silvi-cola, auri-fex, signt-fer, fati-dicus ; corni-ger, arci-tenens, lucti-ficus ; muni-ceps, sorti-legus ; parrl-cida, luc-I-fer, rur-i-cola ; (2) horri-sonus, terri-ficus ; miseri-cors ; (3) undi-que, indi-dem, sici-ne ... hici-ne .... Ante, bene, male vary : anti-cipo, anti-stes ; but ante-cedo, ante-venio ... beni-gnus, beni-volus ; but also bene-volus ... mali-gnus, mali-ficus ; but also male-ficus .... E-verbs compounded with die ere, facere keep e or weaken it to e : valedicere, arefactus, tepefactus. 3. The Suffixes I0-, ro-, cpo-, bro-, bulo-, tro-, tilo-, &c., often change their vowel into 1 before the Nom. ending s ; thus causing Adjectives in us, a, um to pass into the I-declension. gracil-us, gracil-i-s ; hilar-u-s, hilar-i-s. steril-u-s, steril-i-s ; indecor-u-s, indec6r-I-s. seques-ter -tra -trum ; seques-ter -tns -tre. On this preference of 1 the Adjectival forms in li-s, ri-s, cri-s^ bri-s, bili-s, tri-s are founded. By the passing also of 6- ii-s into 1- i-s arises a double form of numerous Adjectives : imberb-u-s, imberb-i-s ; unanim-u-s, unanim-i-s decliv-u-s, decliv-i-s ; effren-u-s, effren-i-s. In bicorni-s, u of the stem passes into 1. See § 28. 4. Before the Ending -bus of Dat. Abl. PI. we have 1 for o, in qui -bus, hi- bus (Plant), and other old forms. I in I-nouns, as navi-bus (nave-bos on the Duellian Column). \ vincular in Cons. Nouns, as reg-i-bus, virgin-i-bus. I foru generally in U-nouns, as canti-bus, corni-bus; except those in -cu-s, and artus, partus, tribus; which keep il. C) The last-cited examples point to that middle sound between \ and u, which the Emperor Claudius wished to mark by a distinct sign. See p. 9. This exists almost exclusively before labials, affect- ing chiefly such words as the following : — (i) imo- or iimo- : lacrima lacruma victima victuma aesttmo aestumo existimo existumo legitimus legitumus maritimus maritumus maximus maxumus decimus decumus monimentum monumentum testimonium testumonium. (2) ip- or up-, lb- or iib- : mancipium mancupium recipero recupero libet lubet ritibus ritubus. (3) If- or iif- aurifex aurufex pontifex pontufex manifestus manufestus sacrifico sacrufico 1 32 Latin Soimdlore. Also capi talis or caputalis and a few more words. Inscriptions shew that the forms with u, prevailed in E. L. and R. L., those with i in and after the Augustan age, for which the Monument of Ancyra, as edited by Mommsen, is the best authority. V Recapitulation, The principles thus laid down respecting the ad- aptation of certain vowels to certain consonants in Latin are sup- ported by the usage of other Italian dialects so far as known. See Corssen, II. 60-225. These principles affect short vowels much more than long; suffix vowels more than root vowels ; grave much more than accented vowels. The general results are : A, the strongest vowel, into which none other is changed, is not itself appropriate to any particular consonant, though its natural kinship is to gutturals first, and least to labials. O is appropriate (i) to v, (2) to 1, r. XT is appropriate to 1 and the Labials, z: is appropriate to r. Z is appropriate to the Dentals n, t, d, s. Again : E and u are appropriate to grouped consonants, z: is convenient for final syllables and the end of words. ZS is a convenient letter for the syllable of Reduplication in Verbs. Z is adapted, by its lightness, to link stems with suffixes, and suffixes with each other. IT, e, sometimes take its place. All these appropriations arise from euphonic assimilation, in- tended to make utterance less troublesome. Again : The extensive weakening of Pr. a through o to u and through e to i, is characteristic of Italian language. In L. L. a reaction occurred, by which o and e recovered much of their lost ground, and in modern Italian o very often appears where u stood anciently : often e where Latin had i : motto, mosca^ polvere, sepolcro, fonder romp ere, sono (sum), &c., bevere (bib ere), disse (dixit), senza (sine), verde (viridis), Assimi- xxiii. Vowelchange by Assimilation and Dis- anTDis- similation of Vowels to each other. tion?^' By Assimilation a letter is changed so as to become the same as another, or so as to become more suitable to it. When a letter is changed so as to become unlike another, this change is called Dissimilation. Every such change has euphony for its object. Assimilation may affect adjoining or disjoined letters. It may be Regressive, when the following letter operates to change the preceding : or Progressive, when the former letter operates to change one which follows. I 1 12, Assimilation and Dissimilation of Vowels. 33 1. Assimilation of Vowels. Assimi- lation of A) Assimilation of adjoining Vowels. Vowels, (An adjoining vowel is never assimilated so as to be the same as its neighbour ; but only so as to be suitable to it.) a. Regressive. i) In the conjugation of the Verb-roots !, gOy qut, can^ and their compounds, 1 before a, o, u is changed into e : earn ... eo, eunt; queam queo, queunt. Z before e in their Participles is used rarely : as Nom. S. iens, qui ens, but in the Oblique Cases usually ie becomes eu : euntis queuntis So iendum ... usually passes into eundutn ... . As ie is an admissible combination, it is probable that the order of change was i-ont" 1-ond-j then e-unt-, e-und-, which remained in this old verb after ent- end- had come in generally. Ambio, one of the compounds of eo, is conjugated like audio. 2) The Pronoun-root 1 (is), arid its strengthened compound idem, in the same manner change i to e before a, u: hence we get ea, eam, eum, eo, eos ; eadem^ eandem, eundem, eodem, eosdem, easdem. 3) Deus, dea (for div-us, a, from Pn div), is an assimilation of the same nature. In Nom. P. di (dei) are used; in D. Abl. dis (dels) ; but not dii, diis. But Diana is classical : Dean a L. L. 4) Teate, Teanum, for Tiati- Tiano- O. nausea (j/avc/a) ; cachlea {KoyXiaq)* but 1 remains in pius ... via (veha). Progressive. 1) By the influence of e or of I preceding it, o is prevented from passing into u in the suffix 6I0- ; see p. 22. 2) Substantives in -ia, Decl. i., pass into -ies, Decl. 5 ; avarit-ia avarit-ies ; mater-ia mater-ies.^ 3) In Numeral Adverbs, from Pr. 1-yans, comes -lens (-les) : ^ quot-iens (quot-ies) ; dec-iens (dec-ies). 4) In Verbs the Mood-suffix la becomes le : (es-ia-m) = siem = sim ; (ama-s -ia-m = ama-ie-m = ama-im) amem. c a \ T^f, ^;^^^ Declension is a mere offshoot of the First. The ending a, Decl. i . was ongmal y long as aqnila in old Latin poetry. Hence came Je by asiimilation from 4 and, with addition of Nom. S. Ending s, les : luxuria, luxurie-s. D Latin Sound lore. EsquTliae (aesculus) exilium (exulo) facHis (facul) Quids (Cures) B) Assimilation of disjoined Vowels. (Regressive and complete always in Classical Latin.) 1) u is often assimilated to a subsequent \ : Aemilius (aemulus) consilium (consulo) -cilium (-culere) similis (simul) ^ manlbiae (manubiae) 2) O is assimilated to a subsequent 1 in inquflinus (incolo) | upilio (ploTrokoQ) 3) s is assimilated to a subsequent \ in Duilius (Duel-l-ius, Bellius), Brundisium (Brundesium) , mihi (mehi), tibi (tebi), sibi (sebi) ; nihil (nehil), ntmius {nQ,-m\-\x-s)y2mme astir ed. See C. 11. 366' famtlia {0. fa77tel, whence famul, famulus).* U is assimilated to o in ^ soboles, when written for suboles. O is assimilated to e in bene (bono-) x: is assimilated to u in tiigurium (tegere). And long e to o in socors (secors). DissimU II. Dissimilation of Vowels. Vowels!^ i) It has been shewn that in E* L. and R. L. u, v wefe avoided before u, whence such forms as vivont, avos, servo m, &c., antiquom, suom, &c., continued in use to the Augustan age. Uv was not so much avoided. We find indeed floviom^ conflovont in E. L., but also in R. L., Cluvius, luventius. 2) The concurrence li was avoided in E. L. and R. L. by writing i-ei ; as fili-ei * sons ; ' peti-ei^ &c., ieis and eeis ; also adi-e^e in Senatuscons. de Bacc, but in I. L. this repugnance faded; and we find iis consiliis, &c. on the Monument of Ancyra. In C. L. ii is avoided by writing e for i in anxi-etas, ebri-etas, pi-etas, sati-etas, soci-etas, vari-etas, abi-etis ...,ari-etis pari-etis ... vari-egare, li-en, Ani-en, ali-enus, lani-ena, and in many Proper Names : Cati^enus, Labi-enus. * Few words have been more debated, as to their derivation and consequent orthography, than suspTcio (suspltio)and con vicium(con vTtium), Each'form has good documen- tary evidence in its favour, and perhaps the strongest argument for t is that, while ci often appears in I. L. and L, L. for ti, converse examples are hardly to be found. Yet Corssen is strongly in favour of suspTcio, as an assimilation of a strengthened form s u s p e c i o, and of c o n vT c i u m, as an assimilated form from convocium. Fleckeisen on the other side assumes suspitio from suspicitio, and con vltiumfrom convocitium. Subiudice lis est. There are strong arguments against each view ; but for the present Corssen's seems the less objectionable. § 12. Vowelweakening in Compound Words, 35 It is avoided in the compounds of jacio by casting out one i, and allowing to the other the power of ji. See pp. 10, 38. Peior is perhaps by dissimilation for pid-ior (compare pi^y * injure In the Pronouns is, idem, the forms li, iis were avoided by writ- ing ei, eis : but li, lis were tolerated in Imperial times. • 3) 0-0 was tolef ated in I. L. But c6-6pia becomes copia ; and coptato is in the Lex lulia for co-optato. M. Lucr. v. 342. ing m Com- pounds. xxiv. Vowelweakening iii the Second Member vowe!- fr^ 1 f T 7 1 weaken Compound Words. Composition of words forms either loose or fast Compounds. If the two members are so joined that, although the first is pro- clitically connected with the second, nevertheless they can be se- parated, the compound is loose. Thus Mars-pater is a loose compound ; but becoming Maspiter, it is fast; because the parts are inseparable. In old language compounds are often found in a state of separation : M. Lucr. i. 452. ob vos sacro (Festus) sub vos placo „ facit are (Lucn) per mihi gratum est per mihi placet obsecro vos supplico vos arefacit pergratum est mihi mihi perplacet Such compounds as satisfacere, circumdare, &c., may be considered loose ; while proficere, tradere, &c. are fast. The fast Compounds hitherto cited, Maspiter, proficere, tra- dere, weaken the root-vowel of the second member. But this weakening, though of frequent occurrence, is not universal in fast Compounds. Thus attraho, though a fast Compound, is not weakened. We have now to see what compound words do weaken the second member of the composition. I. d) Numerous words keep their root- vowel a unweakened in the second member of their compounds ; such are most Verbs of Conj. i. : agitare, amare, gravare, vagari; tnatly of Conj. 2. : ardere, iacere, manere, pallere, patere, pavere, valere ; many Nouns : animus, avus, faber, palma, par. Some words, as will be seen, weaken a part of their compounds, but not all: from mandare, commendo, but demando. Likewise some compounds are not weakened in earlier Latin which are weakened later : M. Lucr. ii. 951, 11 35. aspargere, dispargere (Lucr.) ; afterwards aspergere, dispergere. D 2 36 Latifi Soundlore. § 12. b) A is weakened (through o) into u in the second member of some compounds : a. before I : calcare . . con-culco : in- pro-culco. salsus . . insulsus. saltare . . ex-sulto : de- in-sulto. saltum . . de-sultum : as- dis- ex- in- prae- pro- sub-sultum. Note, Salire anciently was weakened by u, dissuluit (Lucr.); but later it took i by assimilation : de-si lio. /3. Before Labials : cap- . . . occupare : nuncupare : aucup- : mancup-. taberna . contubernium. lavere . . diluvies, al- col- il-luv-ies, -ium. y. After qu, by assimilation : quatere. . concutio, de- dis- in* per- suc^cutio -cussi ... quare . . cur (for quor). Before ss : as, assis . decussis : nonussis : centussis. Note, O (from Pr. a) is weakened into u in consul, exsul, praesul, insula, consulo. Long a is weakened into u in the suffix -ugo (-agro) : aerugo, albugo, ferrugo, lanugo. c) £l is weakened into 6 in the second member of many com- ■ pounds : tam . . autem, item. -dam . . idem, itidem quidem, tandem apisci . . indepisci. canere . . oscen, corntcen, fidicen accentus .... bacillus. . imbecillus ... gradi . . aggredior ... con- de^ di* e- in- prae- pro* trans- re-gredior : aggressus .... lacere . . illecebrae, illectus, paelex. pacisci . . depecisci (or depac-) : but compacisci. pati . . . perpetior, perpessus. fatigare. . defetigo (or defat-). fatisci . . defetisci, defessus. dare. . . addere, de- e- pro- red- tra-dere ... . (Sk. dhoi) . abdere, con- abscon- in- sub- ere- ven-dere. parare . . (impero ... ; pauper propero, aequipero, vitu- pero ...?) but apparo, com- prae- re- se-paro. pario . . comperio, reperio : (aperio, operio ?) puerpera, vipera .... ager . . . peregre (i), peregrinus ; but peragrare. arma . . inermis. arcere . . coerceo, exerceo .... ars . . . iners, sollers, quinquertium. agere . . remex. annus . . biennis, biennium, tri- dec-ennis -ennium aptus . . ineptus ; adeptus. 12. Vowelweakenmg in Compotmd Words, 37 as, assis . tressis, bessis, bicessis .... barba . . imberbis. candere . accendo, incendo ... succendo .... cantus . . accentus, concentus. capere . . particeps, princeps ... auceps, manceps ... . captus , . acceptus, con- de- ex- in- prae- re- sus-ceptus...... caput . . anceps, biceps, triceps, centiceps, praeceps ... .. carpere . . discerpere, con- de- ex-cerpere. castus . . incestus. damnare . condemnare ; indemnatus, indemnis., facere . . artifex, opifex, carnifex. factus , . affectus ... con- de- ef- in- prae- re- suf-fectus but labefactus ... with many more, fallere , . refello. fassus , . confessus, dif- pro-fessus. farcire . . confercio, confertus, infercio, refercio, refertus* iacere . . obex (for ob-iex). iactus . . adiectus, con- de- dis- in- ob- re- sub-iectus lactare . . delecto, oblecto. mandare . commendo, but demando. pandere . dispendo, dispessus (but expando). parcere. . comperco, compesco, dispesco (but com-parsit). pars . . . expers, impertio, dispertio, bi- tri-pertitus (-par- titus). partus . . compertus, repertus (apertus, opertus). passus . . perpessus. patrare . . impetro, perpetro. raptus . . abreptus, eor- di- sur-reptus. sacrare . . consecro, ob- ex- re-secro (consacro, Men. Anc). scandere . ascendo, conscendo, de- ex-scendo. spargere . aspergo, con- di- in- re-spergo. See p. 35. stare . . antistes, supers tes (-stit-). tractare . contrecto, de- ob-trecto ; but retracto (con- tracto, Lucr.). Note, O (Pr. a) is weakened into e in potis . . hospes, sospes (pit-) ... but compos, impos. Long a is weakened into e in halare . . anhelo (redhalo, Lucr. vi. 523). d) IL is weakened (through e) to 1 in the second member of many compounds : agere . . adigo, ab- ex- red- sub-Tgo (but circumago, perago, satago), nav-ig-o. Part. P. -actus, apisci . . adipiscor, indipiscor. amicus. . inimicus.... cadere . . accido, con- de- ex- in- oc- re-cido decMuus, occiduus, . . . stilicidium. canere . • accmo,concmo,prae-pro- re- suc-cino,vaticTnium,^ luscinia .... caput . . occiput, sinciput, ancipit- praecipit- .... capere . , accipio, con- de- ex- in- per- prae- re- sus-clpio^ ... praeclpuus, princfpium ... . datus . , addltus ... de- prae- pro- red- tra-dJtus. 38 Latin Soundlore. Pr. dha . . abdttus, con- e- sub- ere- ven-dttus. facere . . afficio, con- de- ef- in- of- prae- pro- re- suf-Hcio ; cpp. with -ficus -ficium, beneficus ... beneficium ...,but benefacio, calefacio, and all similar cpp. difficilis. superficies, inficetus. conf iteor, dif- pro-f iteor, infitiae, inHtior. adhibeo, co- ex- in- per- pro- red-hibeo; but post-habeo. adicio, ab- con- e- pro- re- in- ob- sub-icio. On forms in MSS. with e, and on dissice, see M. Lucr. ii. 951. alllcio, e- il- pel-licio. delttesco. comminus, emlnus. enim, etenim. luppiter, Diespiter, Maspiter. displiceo : but perplaceo. abripio, arrlpio, cor- de- di- e- prae- pro- sur-ripio. irritus. adsilio, de- ex- in- pro- re- sub-silio. deslpio, inslpiens ; resTpisco. institor, iustltium, solstltium. constituo, de- in- prae- pro- re- sub-stttuo. destino, obstino, praestino, obstinatus. efifringo, in- con- per- re-fringo. Part. P. -fractus. compingo, impingo. Part. P. -pactus. attingo, con- per-tingo. Part. P. -tactus. praefisclne (i). semis, semisses. Note, o (Pr. a) is weakened to X in potis. . . hospita, sospita, hospttium ... . Long a is weakened into i in the suffix -ig:o (-agro) : fuligo, robigo, uligo, &c. 2. d) s is kept in the second member of many compounds : edo, fremo, gemo, meto, peto, seco, sequor, tremo, tego, veho, venio, gen-, ped- ; and those with er, fero, gero, sero, tero. 1: is weakened into i in the second member of several com- pounds : indtgeo, indigus. adtoo, eximo, per- red-imo, (but coemo). coUigo, de- di- e- se-ligo. But intell^go, neglego, sublego. Also perlego, prae- re-lego from legere, to read, dimidius. comprlmo, de- im- op- re- sup-prtoo. ^ arrigo, cor- de- e-rigo. assldeo, con- de- dis- in- ob- prae- re- sub-sldeo ; asslduus, praesidium, subsidium. ' facilis facies facetus fateri habere iacere lacere latere manus nam . pater placere rapere ratus salire sapere stare statuere (stan-) . frangere pangere tangere . fascinare as, assis egere emere legere medius . premere regere . sedere . § 12. Vowelweakening in Compound Words, 39 tenere • . abstlfneo, attineo, con- de- dis- ob- re- per-t!neo ; continuus, pertinax, protmus, protinam. dedi . . addidi, &c. steti . . . adstiti, &c. In close syllables compounds resume e : ademptus, coUectus, compressus, directus, consessuSj retentus. Long e is weakened into i in lenire ^ . delinio (also delenio). tela . . . subtilis. c) £ is changed to u in temnere . contumelia (contumax ?) 3. o is kept in the second member of compounds generally: convoco, abrodo. But locus , , ilico. gnotus . . agnitus, cognYtus. 4. IT is kept in the second member of compounds : ac- incubo, eluceo ; except that u is weakened into e in iurare . . de-iero, pe-iero. 5. The diphthong ae is often kept, as exaest'uo, obaeratus; but melts into i in aequus . . inlquus. aestumare . existimo. caedere . , abscldo, accTdo, con- de- in- oc- prae- sue- re-cido, homicidium, parriclda .... laedere . . allido, col- il-lido. quaerere . acquiro, anquiro, con- dis- in- per- re-quiro, inqui- sitio .... 6. The diphthong oe (ol) sinks to 1 in coenum^ . inquinare, coinquinare. In E. L. it sank to u in ludere, iiti, munus, munio, punio, etc., and their compounds. See xii. 7. The diphthong au is generally kept : inauro, adaugeo: but it sinks to 0 in faux . . . suffocare ; plaudere . explodo, supplodo (but applaudo) ; to u in causa . . acciiso, incuso, recuso ; fraus , . (frustra, frustrare) defrudare : see M. Lucr. vi,.i87^ claudere . concludo, dis- ex- in- oc- prae- re-cludo ; and to oe in audire . . oboedire. Note, The other Italian dialects exhibit the same general laws of Vowelchange as the Latin. * Obscenus(obscoenus) is usually derived from c o e n u m. This, however, is by no means certain. 40 Latm Soundlore. § 12. Redupii. XXV. Reduplication, cation. Reduplication in language is a practice as old as language itself. The infant from instinct or imitation forms words by repeating the syllables : pa-pa, ma-ma, ta-ta ; often unconsciously weakening the first : pu-pa, me-ma, ti-ta : and the mother or nurse amuses or lulls the infant by similar repetitions : ding-dong, by-bye, &c. Various emotions express themselves in the same manner : aha oho ! &c. See Pott {Die Doppelitng). Thus arose the habit of modifying words A) By doubling a root merely: B) By prefixing to it its first consonant and vowel. After which it came to pass, that the reduplicative syllable might be either strengthened or weakened, and the root itself weakened (rarely strengthened) after reduplication, in consequence of accentual change. A) Reduplication by doubling the Root merely : d) bar-bar-US (bulbul Pers.), cu-cu-lus, la-la-re, Mar-mar, cin- cin-nus, tin-tin-nare, ul-ul-are, cur-cul-io, gur-gul-io, fur- fur, mur-mur, tur-tur. So quisquis, utut, ubiubi, &c. b) The Root is weakened in car-cer, mar-mor. E) Reduplication by prefixing the first two letters of the Root. (This is specially important in Greek and Latin on account of its use in forming the Perfect Tense of Verbs.) d) Without vowelchange : cd-cul-lu-s, {po-pol-ti-s), su-sur-ru-s, and the following Per- fects ; cu-curr-i, d!-dic-i, mo-mord-i, pe-pend-i, p6-posc-i, pu-pug-i (pu-;/-go), scT-cid-i (sci-;/-do), spo-pond-i (spon- deo), te-tend-i, to-tond-i, tii-tud-i. b) Redupl. weakened, Root unchanged ; in occasional forms ce-curr-i, me-mord-i, pe-posc-i, pe-pug-i, spe-pond-i. c) Redupl. unchanged ; Root strengthened. pa-pa-ver, tu-tud-i (rare). d) Redupl. strengthened ; Root weakened. Ma-mers, Ma-mer-cus, Ma-mur-iu-s, pa-pll-io, p5-pul-ug {poplar)y pu-bl-icu-s. e) Redupl. unchanged ; Root weakened^ p6-pul-us {people), /) Redupl. and Root weakened. cT-cind-ela (candela), ci-con-ia ; tl-tu-lu-s ; b!-be-re (po Pr. pd, drink)y gi-gn-ere (Pr. gdn, gen, engender)^ si-ste-re (sta-), se-re-re (for se-se-re. Root sd). The reduplicative syllable is weakened in many Perfects by changing its vowel to 6 (see xxi.) : § 12. Assimilation of Consonants, At de-d-i (da-), ste-t-i (sta-) : fe-fell-i (fallo), pe-per-i (pario), pe-perc-i (parco) : te-tul-i (tol-l-o, Pr. tal) : ce-cid-i (cado), ce-cin-i (cano), pe-pig-i (pa-;/-go), te-tig-i {tdi-n- go) : ce-cid-i (caedo). Obs. A consonant is lost in si-ste-re (for sti-ste-re), sci-cid-i, usually scid-i (for sci-scid-i), sp6-pond-i or spe-pond-i (for spo- spond-i or spe-spond-i), p6-pul-are (for spo-spul-are from spolium). A vowel is lost in de-d-i (for de-de-i) : gi-gn-o (for gi-gen-o). A vowel and consonant are lost in ste-t-i (for ste-ste-i). xxvi. Changes of Concurrent Consonants. (The sign x is used to express ' becomes/) I. Complete Assimilation of Consonants. A) Regressive Assimilation : Assimi- lation of Conson- ants. (dq) X cq (adquiro) acquiro „ (quidque) quicque (dl) X 11 (adludo) alludo (sed-^-la) sella (conloco) coUoco (coron-//-la) corolla (un-^-lus) ullus (perlicio) pellicio (ager-//-lus) agellus (tn) X nn (pet-na) penna (dn) „ „ (adnuo) annuo )9 J7 V (bm) X mm (submoveo) summoveo (sub-mus) summus (flagma) flamma (inmotus) immotus (subripio) surripio (inrideo) irrideo (&m) ,y (nm),, (bp) X )9 rr (as) 3> J> X SS (ts) 3> V yy (merced-narius) mercennarius. (fod-sa) fossa (adsurgo) assurgo (cedsi) cessi (concutsi) concussi The following Assimilations also occur in the Composition of Particles with Verbs : (be) X CO (obcurro) occurro (dc) „ „ (adcedo) accedo (b&) X (obgero) oggero (dgr) „ „ (adgravo) aggravo (dt) X tt (adtendo) attendo (bp) X pp (obpono) oppono (dp) „ „ (adpeto) appeto (bf ) X ff (obfero) oflfero (cf ) „ „ (ecfugio) effugio (df ) „ „ (adficio) afficio (sf) „ „ (disfiteor) diffiteor a) (nd) X nn occurs in Plautus : dispennite for dispendite ; distennite for distendite. So in Oscan ; opsannam = op Gra.n dam, (3) mn, though stable in C. L., often yields to assimilation in modern language : L. column a, It. colonna, Fr. colonne, y) That final m of a proclitic word assimilated itself in utter- ance to a following n, is testified by Cic. Or. 45 and Quint, viii. 3. 45. Thus etiam nunc was sounded etian-nunc.^ ' The sharpening of an inner syllable by doubling a consonant (relligio,reIliquiae, millia, querella, bracchium, luppiter, littera) must not be confounded with Assimilation. See Appendix A. : also C. I. 227. II. 466. 42 Latin Soundlore, B) Progressive Assimilation : (ferse) ferre ; (farsis) farris ; (rvpaig) turris. So C. forms (miser-timus x miser-simus) miserrimus. (is) X 11 ; (vol-se) velle ; (mel-tis x mel-sis) mellis, &c. (facil-timus X facil-simus) facillimus (C.). (st) X ss : (duris-timus) durissimus : where duris is contracted from durius (C.).^ (This assimilation occurs in some Supines, according to C.'s view: fissum, fossum, passum, &c. See xxxi.) II. Partial Assimilation of Consonants (Adaptation). I. The Sonant g becomes c, and the Sonant b becomes p, before s or t : (reg-si) x rexi ( = rec-si) (scrib-si) x scripsi (reg-tum) x rectum (scrib-tum) x scriptum a) But ab, sub, ob, may remain in composition : absens, subter, obtineo (but also apsens, optineo) And bs final may be kept in Nouns : caelebs, plebs, trabs, urbs (but also pleps, urps, &c.). Ods. X = cs = any Guttural with s : any Guttural except c being supposed to become c before and so to form x : ' (dic-si) X dixi (sug-si x suc-si) x suxi (fac-s) X fax (leg-s x lec-s) x lex (coqu-si X coc-si) x coxi (ungu-si x unc-si) x unxi The following Verbs deserve special attention : trah-ere Perf. (trah-si trac-si) traxi (from a lost Pr. tragk f) veh-ere — (veh-si vec-si) vexi : Sk. vah (a lost Pr. vagh f), * The formation of Latin Comparatives and Superlatives may be briefly stated here. I. Comparatives. 1) (Sk. ydnSy yas) Lat. (-ios) X -ior -ius is added to the Ch'pt Stem : (dur-ios) dur-ior, dur-ius ; (ingent-ios) ingent-ior, ingent-ius. (mag-ios, mag-ior, &c.), ma-ior, ma-ius : — magis for mag-ius. (root min- ; min-ior, &c.), min-or, minus, (root pie- =Sk. : ple-ior, ple-ius, plo-ius, pious), plus, plur-, 2) (Sk. tard) Lat. tero- is added to Roots and Stems : al-ter, u-ter, dex-ter, sinis-ter, in-ter : — pari-ter, ali-ter, &c. 3) Both Suffixes are used in mag-is-ter, min-is-ter ;— dex-ter-ior, in-ter-ior, &:c. IL Superlatives. 1) (Sk. ta)X.o-y inquar-tu-s, quin-tu-s, quo-tu-s, &c. 2) (Sk. ma) mo-, in sum-mu-s, i-mu-s, pri-mus, mini-mu-s, pluri-mu-s ; (ex- ter-mu-s)xextre-mu-s ; (pos-ter-mu-s)x postre-mu-s ; (super-mu-s) X supre-mu-s. 3) (Sk. tamo) timo- in ci-timu-s, ul-timu-s, op-timu-s, in-timu-s, ex- timu-s, pos-tumu-s, dex-timu-s, sinis-timu-s. passes into simo- in (pe-d-timo-) pessi-mu-s, (mag-timo-) maxim-us, proximu-s. passes into (simo-) ITmo- in facillimu-s, &c. — — (simo-) rimo- in miserrimu-s, &c. In most Adjectives tlmo- is added to the contracted comparative is (iros) and as- similated : (dur-is-timo-)x durissimu-s. So tristissimus, felicissim,us, &c. g 12. Dissimilation of Conso7iants. 43 viv-ere Perf. (vigv-si vic-si) vixi : Sk. jiv (Pr. gvigv-). flu-ere — (flugv-si, fluc-si) fluxi (from a lost form flug-vere), stru-ere — (stru-ic-si) struxi (probably from 2iiQxmstru4c-ere), Add the nasalized ninguere with its Noun nix, s-now (Pr. snighj Sk. snih^ ' to stick'). Nix (ningv-s) drops v in Nom. Sing, and in the other cases, forming Gen. niv-is, &c. 2. Liquids and Nasals ^ take Sonants before them in preference to Surds : (po-pl-icus) X pu-bl-icus (ili-cn-us) x ili-gn-us (qua-tr-a) x qua-dr-a (cy-cn-us) x cy-gn-us (ne-cl-ego) x ne-gl-ego (se-cm-entum) x se-gm-entum Through some feeling of euphony (nec-otium) becomes neg- otium. 3. TX becomes m before the Labials m ; but remains be- fore f, V : impleo, imbuo, immitto ; but infero, inveho. 4. A Labial Mute becomes m before n : (sop-nus) X somnus ; (Sab-nium) x Samnium. 5. iw: often becomes n within words before a Guttural or Dental ; and, if kept, is sounded as n : clan-culum prin-ceps eun-dem clan-destinus quen-dam ean-dem. So quon-iam for quom-iam. But in some instances m must be kept : quemque, quemquam, unumquemque, namque, numquis. In others m is better than n : quamquam, tamquam, cumque, umquam, numquam. 6. When Dental Mutes meet, the former often becomes s : (edit, ed-t) x est (claud-trum) x claustrum (rod-trum) x rostrum (plod-trum) x plostrum. In Supines and Superlatives sometimes both become s : (fod-tum) fossum ; (pat-tum) passum ; (duris-timus) durissimus. III. Dissimilation of Consonants. Dissimi- lation of The recurring sound of the same Consonant in succeeding syllables is sometimes avoided by changing it in one place. a) caeluleus, caelulus are changed into caeruleus, cae- rulus. b) Palilia is sometimes written Parilia : Remuria x Lemuria. * The assimilation of Sonant to Nasal explains the sound of g"!! in French -grne final, as cygne. Its sound in Frenchand Italian before interior vowels=n-y ; thus, agneau, agnello (=an-yo, an-yello). 44 LatUi Sotmdlore, § It. Trans- posi- tion. Eu- phonic Inser- tion. 1.055 of Initial Letters. c) The suffixes all- ell- ill- ull- are chosen for Adjectives de- rived from Nouns, if the root contains p : and the suffix -arl is chosen if the root contains l ; austr-ah-s cardin-aH-s liber-ah-s reg-ali-s crud-eU-s puer-iH-s cur-uli-s al-ari-s capill-ari-s coll-ari-s sol-ari-s stell-ari-s tutel-ari-s vulg-ari-s Obs, But in the suffix -ario, p is not changed : ordin-ariu-s, temer-ariu-s. Note I. Consonants are sometimes transposed within a word for the sake of euphony : pristis for (pistris) I columns for (corulnus) extremus „ (extermus) | Note 2. When m is followed by « or t, p is euphonically inserted to strengthen the syllable : hiem-p-s, em-p-tor, sum-p-si, sum-p-tum.* The change temfitare for ten tare, though supported by inscrip- tions and good MSS., is censured by Corssen as an etymological blunder : the formation of the Verb being Pr. tan, L. ten, whence ten-d-ere, ten-tu-s, ten-t-are. s seems to be euphonically inserted in mon-s-trum, mon-s-tro, &c. (from mon-eo). (On the euphonic insertion of a Vowel in m-t-na, drac-u-ma, &c., see xxii. On the insertion of e in ag-e-r, nig-e-r, &c., see xxi.) * The Loss of Letters will next be considered. g xxvii. Loss of Initial Letters {'Acfyaipsaci), lost by lamentum... laus; luscinia ... vapor ... nasci, natus ... noscere, notus ... narrare lac shewn in c-lamare c-luere g-nasci g-natus : Sk. jan Gr. yei/- g-noscere, i-gnotus : Sk. jnd Gr. (g-narigare) from g-narus Gr. ya-Xa/c-r- * The euphonic insertion of b between m and 1 or r, and that of d between n and r, occur in Greek (as /Lte>-i3-Aw). Some think that ^2/ should always be assumed as the primitive of qu ; but Corssen maintains that c (k) could develop u after it in Italian language as a transition-step to the labial p : and he thinks that even in Indie kv is developed from k. 3) The Labialism by which tt and p represent Pr. y^, prevailed in Umbrian and Oscan. U. /^///^^^//^^^^y-quadrupedibus ; O. ^///// = quidquid. Hence (from O.petora, four) come the names Petreius, Petronius: and (from O. /^;;^-/-^j' = quinque) Pon- tius (=Quinctius), Pompeius, Pompeii, Pomponius, Pom- pilius.^ 4) In two instances c, qv seem to correspond to Sk. Gr. tt : L. Sk. Gr. coquo pac' TrcV-roi quinque pane' a ireyrej TrifiTTE, Here some think the primitive roots were kak or kvak; kanka kvankva, Fick, however, supposes coquo to be for (poquo), quinque for (pinque), by assimilation. 5) In proof that qu could be developed from c, Corssen cites huiusque for huiusce ; inquilinus from incolo ; inquinare fromcoenum; quom = cum; querquetum for quercetum ; Qui- rites from Cures; sterquilinium from stercus. * Perhaps other instances of Labialism (p for k) in Latin are dialectic (Sabine) : as lupus {^\i.var-kas, Gr. Avko?) ; popina for coquina ; palumbes = columba : Epona (forEquona) ; spoiium (Gr. aKvKov) : and one or two more doubtful, as pa vo (Gr. raw?). Of Dentalism in Latin (t for k) the traces are few and dubious : as talpa (for s-talpa, /S-calpa) : stercus (Gr. o-Koip) ; studeo, studium(Gr. anov^ri). 6o Latin Soundlore. ■ ■ § 12^ 6) Q is found in E. L. for qu, chiefly before u, as peqimia^ qum, qur^ quius : rarely before other letters : as neqidem, qe, 7) To avoid quu, before the Aug. age on was often used : ecus, cocus, anticus, execuntur, secundus. So locutus, secutus. Ne-cubi, si-ciibi, ali-ciabi, &c., take the place of ne-quubi, &c. 8) Qu becomes in Greek fcov, fcv^ kt, sometimes ko : 9) Qu was uttered as in English ; c as k.* Their sounds appear in ' come quicker.^ The assibilation or soft sound of cl, ce did not prevail in Latin before the 7th century of our era. * I. The following facts shew that the assibiUted sound of ce,,ci, wa^ not used in C. L. 1. Greek represented c by k before e, tj, t : as 2. Latin represented Greek k by c before e> i» y : as Cecrops, cerasus, Cilix, Cimon, cithara, Cybelc. 3. Gothic represents c by k before these vowels : as kerker, keller, kirsche. 4. Quintilian cites chenturiones as a way of spelling centui^loiies. An Inscription a.d. 326 gives schenicos for scenicos,. and also sceniconim. Another, a.d. 408, has p ache for pace. 5. Qu could not represent an assibilated c ; therefore such forms on Inscrr. in L. B.. as {on the one hand) h u I u s q u e for huiusce ; r e q u i.e s q u e t for requiescit and (on the other) sic is for siquis ; eintus for quintus. shew that up to their date ce, ci kept the hard guttural sound. 6. In the imitative verbs crocio, glocio, c must have had the hard sound. 7. Finally, no grammarian has told us that c was uttered in one way before e,. i> in another before the remaining vowels. This silence goes to prove that no such difference existed in C. L. In the Umbrian and Volscian dialects there had existed a soft 9, as U. fafid^^ V. j^jzVi;, for facia t. ' And in the late Imperial times such tendency dawns in a few words on Inscrr. : provinsia for provincia; Luziae for Luciae ; Felissiosa for Feliciosa. But it was not until the 7th century a. d. that popular utterance so far relaxed its energy as to adapt generally the guttural consonant to the palatal vowel, and propagate that sibilant sound of ce, ci which, for instance, transmutes the classical Kikero into It. G. Fr. Eng. Chichero Shishero Sisero. II. The assibilation of inner ti before a vowel began earlier. It had existed in dialects : as U. purdinfMst for purdintiust : O. Bansae for Bantiae. The grammarian Pompeius- in the 5th century testifies that Titius, for instance, was sounded Titsius, Consentius says that etiam was pronounced eziam. In the next century we meet with a.K7^io for actio., Constanzo for Constantio : soon after with iustizia, milizia, preparing the way for modern. Italian, which writes Firenze (Florentia), Piacenza (FlAcentisi), /alazza (palatium)^ 1 Relations of Consonants^ etc. 6i IL The Guttural Sonant C. 1) G usually corresponds to Sk. J or gr, Gr. 7 : G, L. Sk. Gr. gen- jan ytv- yov- ag-o aj ay- teg-o sthag oTiy-fa Sometimes to Sk. s\ Gr. k : viginti vins'ati feiKOcru Sometimes to Sk. kk, k, Gr. x^y- unguis nakhas orvj(- li(n)go Izk Xei)(u} ego aham lyw 2) Parasitic u follows gr in anguis, sanguis, unguis, lingua, linquo, stinguo, tinguo, unguo, urgueo. In pinguis (ttq- yyc) u is a suffix. In all these, except urgueo, the guttural n adulterinum strengthens gr, giving it a nasal twang : as in the Verbs cited p. 19. 3) G was guttural in C. L. ; as in Eng. go, gave, give, get, begety begin. Its palatal assibilations before e, i, whether hard, as in Eng. gentle, giant, rage, It. gentil, Ginevra, gioia, ragione, or soft, as in Fr. gentil, geant, gite, rage, began towards the 5th century with the use, as in Italian, of gri ( = Eng. j) before another vowel : Giove, Giulia, giallo, III. The Aspirates : h, f. 4tet It belongs not Only to Indie language but also to Greek to aspirate the medial mutes gr, d, b, as well as the tenues k, t, p. Thus arise the medial aspirates grli, dli, bli ; to which the partially corresponding sounds in Greek are 0> severally. Latin has neither class of aspirates : the letters which it uses for the purpose of correspondence are principally h, f, and the medial I. K, when sounded at all, was sounded as the Greek Rough H. Breathing, but corresponds to it only in words borrowed from the Greek : Hebe, Homerus, hora, &c. d) In some words li corresponds to Sk. h, Gr. : as L. Sk. Gr. hiemps M^nam X^'M" heri hyas y^ic, veho (via) vah ^X^^ ans-er(for ^ X^>, h-ans-er) - ( = * The Teutonic names of this bird, goose, gander, gos (Anglo-Sax.), gans (Germ.), com- pared with the Greek xr^v, seem to shew that ghans is the Prim. form. There can be no doubt that Greek x indicates a Prim, gh ia all these words ; and this is also shewn in the Latin Perf. of veho : vexi for vegh-si. 62 Latia Soundlore. % 12. U) B represents dialectic fin some words, as haedus, harios* lus, hircus, hordeum, horreum, hostis, also in mihi. So in Spanish, /^//'^ = filius ; /ladlar = {3,hu\3.r'i. c) H has no position in Latin metre ; and a tendency to get rid of this aspirate, as a troublesome sound, is manifest in the history of Latin. Hence the fluctuation in the orthography of many words in MSS. and Inscrr. : harena, arena; harundo, arundo; haruspex, aruspex ; have, ave; haedus (aedus) ; ha.Y \ o\u s (arwha) ; Ha- dria {Adria); heres (eres); hcrus, hera, and erus, eia;hedera (edera) ; ho lus (olus) ; Hammon (Ammon) ; Hister (Ister). But the forms humerus, humor ^ &c. for umerus (^/xoy), umor (from uvere), are not good. c) The loss of li was propagated in L. L. Hence in modern Italian it is not sounded, and has generally disappeared as an initial letter. F. 2. The Italian Labiodental Aspirate f is described by Quintilian as a very strong rough sound : * 111a quae est sexta nostrarum paene non humana voce vel omnino non voce potius inter discrimina den- tium efflanda est/ xii. lo. This description does not seem to imply that the ancient pronunciation of f was materially different from our own : but it does imply what is probable on other grounds, that 0 was different from our f, not, like this, labio-dental, but a pure labial aspirated. T is seldom the inner letter of a root. As an initial it corresponds to Sk. bh^ Gr. (^, chiefly : Sk. dh, Gr. 0, some- times ; Sk. gh^ Gr. rarely. L. Sk. Gr. l) fero bhar (p/pa, fui (fe-, &c.) bhu (f)yu} flag- (fulg-) bhraj 4^\ty(a frigo bhrajj ^puyw fugio bhuj (jifvyto frater bhrdtar (j^paTrip See bhay p. i6. To Pr. bh, C. also refers the f in tnaiiy words : fovere, favilla; favere ... ; famulus (O, /aamay 'house'); fervere furere ... ; fidere ... ; fiber; forare ; furvus; fundus...; frequens: compare also fagus ((/)*?7oc); folium (^uWoj^); frango (ppay-) ; frigus (fplyoQ), 2) foris dvdr Ovpa fumus dhumas Ovog rufus (rudh) epvOpog firmus dhar (da\- deX'Y From this last root C. deduces a large number of words : fere, ferme, frenum, forum, furca, fulcio, &c. * Lat. -fendo, Gr. detVo) are referred to Sk. Ann. Probably on this account Prof. Monier Williams, in his Lexicon, refers Aau to a Prim. dAau, though so many of its forms indicate an original £Aan. § 12. Relations of Consonants y etc. 63 The Preposition af which appears in Latin Inscrr. is by Cors- sen distinguished from ab,and derived from Sk. adhu L. Sk. Gr. 3) fel (comp. bills) {hari^ ^ greenish yellow ^6Kr) fu(;^)do Ig^^f) x^f^ To Pr. grh C. refers fulvus (helvus), host is (fostis), hario- lus (^inspector of the hira or entratP), haedus (faedus), hordeum (fordeum), fames, far, frio, furfur. IV. The Labial Mutes p, b. I^abials P. B. These were sounded anciently as in modern language. P corresponds generally to Pr. Gr. tt. But see L S corresponds often to Pr. Gr. /3 ; but, as already shewn (L and I XL), it has several other special relations. Thus it is developed not only from dv (as in bis, bellum, bo- nus, see p. 45), but also from gv : L. Sk. Gr. bos, bov- gaus pov$ faba {d/ias, ^ eat ') ^ay* (for fag-va) As an inner Consonant b represents Sk. dk regularly, dk rarely. L. Sk. Gr. nubes nahhas vecj^os uber iidhar ov6ap So b = <3f) (M) in am bo (aficjico), ambi- (dficjit), glaber (yXa^vp- off)) nubo (j/v/Lt» volo i)ar /3ou\- ulna aratni Ci\kvr\ sal sara a\c levis iraghus ^kayiq See the derivatives of svar^ p. 17. Lat r from Sk. /is shewn in rump o from lup (old form rup\ 3) Comparing Latin and Greek, we find, on the one hand, lacer (joafcoe), lilium (Xc/piov) : so luscinia (Fr. rossignol) : to the other, grando (xaXa^ct), hirundo (x^Xl^ujv), arx {aXfc-), vermis (eXpc), strigilis (orXeyytc) : with a great number of v^ords in which the letters correspond, especially those with 1 : as leo {\eu)v), levis (Xeioo), oleum (eXaiov), silva (vXrf), 8cc. but also some with r : as aranea (dpaxvrj), rivus (pooc), taurus (ravpoy), &C, Relations of^ Consonants^ etc. 4) Ii\ Latin words the order Mute-Vowel- Liquid often appears where the corresponding Greek forms have Mute-Liquid-Vowel : bardas (Ppadvs), caro (/cp^as), cerno {Kpivco), dulcis (yXvKvs), pulmo (irX^vfiodv), sorbeo (/^o(/)ecD), torqueo (rpeTro)), So tri and ter, trinus and ternus, porro for (protro), &c. 5) Frequent interchange is found between the Liquids and the Dental d : d and 1 : lacrima {bakpv, tear), lingua (E.L. dingua^ * tongue'), levir (Sk. devar, Gr. 6a7;p), olere {obcoba, odor), Ulixes (Obva-a-evs), adeps (aXei^o)). Meditor (/ieX^raoj) is not so certain. d and r : meridies for (medi-dies) ; and ar- for ad in bid compounds: arbiter (ad-bitere), arcesso for (ad-ci-esso). 6) As to the sound of 1, we leaM from Priscian the opinion of' the elder Pliny : ' 1» triplicem, ut Plinio videtur, sonum habet t exilem, qiiando geminatur secundo loco posita, ut ille, Metellus ; plenum qxiando finit nomina vel syllabas, et quando aliquam habet ante se eadem syllaba consonantem, ut sol, silva^ flavus^ clarus; medium in aliis, ut lectum, lectus,' L 7. 38. 7) The lightness of inner 1 caused it to be often shai*peried by doubling : loquella, querella, &:c. 8) Oh its affinity to u, see xx. In French this goeS so far that U often takes the place of 1, forming diphthongs <^//, eau, eu, ou : «(ad illu) X au ; (ad illos) x aux ; (alter) x autre. ; (cheval-s) x chevaux ; (chevel-s) x cheveux. (bel) X beau ; (castellum) x chateau ; (fol, mol, sol) x fou^ moUy soit^ a) No relation is more important in Latin Wordlore than that R and which arose between the letters r and s, changing the sibilant between vowels into the canihe liquid. Varro mentions it : 'In multis verbis in quibus antiqui dicebant s postea dictum r, ut in carmine Saliarium sunt haec : ... foedesum, plusima, meliosem, asenam,' vii. 26. In the Carmen Afvale the Lares are called Lases. Cicero says (Fam. ix. 21) that L. Papirius Crassus was the first to call himself Papirius (B.C. 336) : before which all his clan were called Papisii. So the Auselli became Aurelii, the Fusii Furii, the Numisii Numerii, the Pinasii Pinarii, the Spusii Spurii, the Volesi Valerii, the Vetusii Veturii. Thus we have Halesus, Falisci, and Falerii ; Etrusci, Tusci, and Etruria. Hence in roots these changes appear : (asa)xara; (asena, fasena) xharena; (fesiae) xferiae ; (nases) xnares, comp. nasus ; (geso) x gero ; (hausio) xhaurio ; (seso) xsero; (uso)xuro; (hesi)xheri, comp. ^0fg, hester nus. So spes andspero; quaero and quaeso ; vis, vires; glis, gliris ; flos, floris, &c. ; nefarius from nefas, &c. F 66 Latin Sotindlore, § 12.' Hence almost all the Noun-flexions in r-, as er- op- 6p- ur- from Nominatives in es, is, iis, os (op), us, belong to stems which are really not p-stems, but s-stems : the old forms, many of which are found in old Inscrr., being, for instance {aesis, foedesis, pignosis or pignesis, arbosem, Jioses, plushna, maioses)^ &:c. The Case-endings -apum -opum were {-asum^ -osum). The Verb-forms -eram -epo were {-esam -eso), -pis -pe -p1 were (sis -se -si). In the Passive endings -op -up, &c., p represents the pronoun se. Dir-imo is for dis-emo, dir-ibeo for dis-hibeo. b) The p for s between vowels very often corresponds to the loss of Greek o" between vowels : (ausosa) x aurora (au-wc, Sk. ushas); (ausis) x auris (oJ-a< ); (visus) X virus {fi-oq^ S^.vishas) \ (nusus) x nurus (rvoc, Sk. snushd) ; (sosor) X soror (o-ap, Sk. s^vasar, * sister'); (genesis) x generis (yfVf-oc); (musis) x mu-ris (^v-6c) ; (deasum) x dearum (^^fo-wr); (esam) x eram (e-^i^), &c. c) R is for s before a consonant in Minerva (Sk. ma7tas,^vomd^) ; verna {^\i,vas, 'dwell') ; veternus from vetus, diurnus, hodi- ernus from dies : And as final in the ending op for 6s : color, honof, labor, &c., for colos, honos, labos, &c. S. na) The Greeks, who avoided sibilation as much as possible, sub- stituted generally the rough breathing for primitive c at the begin- ning of words. Not so the Italians. Hence Latin iniiial s before a vowel corresponds often to Sk. s, Gf. aspirate : salix (kXiKY]), sex (e^), sedes (t^oc), semi- (rifJ^t-), serpo (tpTrw), si- mul (a/x«, ofiov), sollus (oA.oc), silva {v\r}), se (t), suus (^fog), suavis (//Sue), sub (vTTo), super {virep), sudor (i^ipujc)^ sus (uc), &c. Sometimes initial s corresponds to Greek ' spiritus lenis : * si (el), sero {^'f-p^o)^ serum (c'ooc). db) Sc, sp, st initial generally correspond in Greek and Latin, unless s is dropt, as in tego (nrfyw). See p. 45. cc) S initial was probably sounded more sharply than as an inner letter : hence caussa as well as causa appears in MSS. and Inscrr., and other occasional doublings of s are found. dd) S falls out in Cerealis for (Ceresalis) ; in ver (i-ap for Feaap) ; in vi-m, v-i ; in the cases of spe-s for (spe-r-es = spe-s-es), in those of dies, die i for (die-s-i), &c., and in other forms. Jhe VI I L The soft Labial Spirant V. Spirant V- d) v-consonans has the same relation to f that b has to p : it corresponds to Pr. v, Gr. digamma, like which it was sounded : and this sound was probably that of Eng. w.* Corssen thinks its * That Latin v-consonans had the sound of Engli^ w always, is probable for the following reasons : i) By a slight change in the position of the speech organs the vowel i passes into y-cons» By a precisely parallel change the vowel u becomes, not Eng. v, but Eng. w. §12. Relations of Consonants, etc. 67 initial sound was that of Eng. v, its inner sound that of Eng. L. Sk. Gr. vomo vam r ' rEJJLEb) volvo val r£AV(ji) voco vac' video vid know') Pt^- vestis vasis novus navas ovis avis So vis (F/c), viola {Ptov), vinum (fotvoc), bos bov- (/3ofc /3ovc), navis (vafc, vavq), ver {Frjp), vespera (fecTrepa, kaTripa)^ Vesta {FecrTia, eor/a), radix {Pptici, jo/^a, /3pt^a), &c. As the Greeks lost the use of they represented initial v sometimes by ov, sometimes by /3, Vafro (Ovappcov or Bdppcov), On the vocalization of v see p. 10 5 on its omission, see pp. 57, 58. IX. X-consonans (J). The On the sound and uses of i-consonans (j), see viii. i, and xii. 6. jf^^' It corresponds to Sk. j/, sometimes to Gr. as iugum (Sk. j^uj] Gr. Cvyov). a) A form of 1 taller than the adjoining letters (I), appears in late Republican and Augustan Inscrr. to express 1) long i-voc. : ^ dIvo, eI, stipendIs* 2) i-cons. ; both between vowels and initially : MAlOR, CViVS, EiVS ; IVS, IVLIA. A more corrupt form il is also found : cviivs, coil VOL 3) That which is merely a general fact, has beeii Wrongly set down as a rule of sound : namely, that a vowel before i-cons. is long. Corssen has shewn that in all words which can be traced (for ieiunus is obscure) where a vowel is long before i*cons., it is so by its own nature : a-io, Ga-ius, Ma-ius, pe-ior, pe-iero, &c. 2) Greek ov (as in OveXia for Velia in Dion. Hal. ) expi'esses Gr. digaiftnla and Lat. v ; and this sound cannot be interpreted as Eng. v, but as w. The occasional substitution of P, by Plutarch chiefly, proves nothing to the contrary : but only means that, ov being a clumsy representation of f and v, /3 was taken as the nearest labial instead. 3) A. Gellius cites a grammarian, who says that Deus Vaticanus presided over infancy, and that the two first letters of his name (Va) are that sound which the infant first utters. The sound then is Eng. wa not v3,, which the infant, having no teeth, cannot utter. Corssen's opinion is that Latin initial v may have had that middle sound between w and v, which German w has in some localities ; the upper teeth being brought near to the lower lip, but not pressed upon it. This view we cannot accept. * It was shewn (p. 33) that ei was long used to express T with a leaning to e. Lucilius tried to mark long vowels by writing them twice, as Maarcus for Marcus. This appears on some Inscrr. but did not last long. It was followed in Cicero's time by the Apex or mark over a long vowel, like (') or ('), which frequently appears on a, e, o, u ; not on i. ¥ 2 68 Latin Soundlore. § ,2, Inbi-iugus, quadrt-iugus, tri-iuges, &c., I, being naturally short, remains so. c) Progressive assimilation has changed 1-cons. to i in cello, fallo, pello for (cel-yo, fal-yo, pel-yo), as a\Aoc in Greek for (aA-yoc), (TCpaWu) for (o'