are inadmissible in the Aristotelian paradigms. A claim, however, is set up for their admission. Grant it, and you may say instead of most—

but this is only a particular instance. You may combine any two numbers in any way you like, provided only that the sum be greater than unity. Now this may be arithmetic, and it may be fact; but it is scarcely formal logic; at any rate it is anything but general.

It is the logic of some and its modifications one, all, and anything between one and all, as opposed to the logic of the simple absolute some (some the opposite to none), and a little consideration will show that it is also the logic of the probable, with its modification the proven, (proven is probable, as all is some,) as opposed to the logic of the possible and impossible. Let, in such a pair of propositions as—

the number expressed by some, as well as the number of the men of the brigade, be known, and the question as to whether

is a problem in the doctrine of chances. One per cent. of each will make it very unlikely that the single brave man was also the single killed one. Forty-nine per cent. of each will make it highly probable that more than one good soldier met his fate. With fifty on one side, and fifty-one on the other, we have one at least. With all (either killed or brave), we have the same; and that without knowing any numbers at all.

III.
GRAMMATICA.

ON THE RECIPROCAL PRONOUNS,
AND
ON THE RECIPROCAL POWER OF THE
REFLECTIVE VERB.

READ
BEFORE THE PHILOLOGICAL SOCIETY,
MARCH 22. 1844.

The present paper is upon the reciprocal pronouns, and upon certain forms of the verb used in a reciprocal sense. It is considered that these points of language have not been put forwards with that prominence and care which their value in the solution of certain problems in philology requires. Too often the terms Reciprocal and Reflective have been made synonymous. How far this is true may be determined by the fact that the middle verbs in the Icelandic language have been called by so great a philologist as Rask reciprocal instead of reflective. This is equivalent to treating sentences like we strike ourselves, and we strike each other, as identical. Yet the language with which Rask was dealing (the Icelandic) was the one of all others wherein the difference in question required to be accurately drawn, and fully pointed out. (See Anvisning till Isländskan, pp. 281, 283.)

In all sentences containing the statement of a reciprocal or mutual action there are in reality two assertions, viz. the assertion that A strikes (or loves) B, and the assertion that B strikes (or loves) A; the action forming one, the reaction another. Hence, if the expression exactly coincided with the fact signified, there would always be two propositions. This, however, is not the habit of language. Hence arises a more compendious form of expression, giving origin to an ellipsis of a peculiar kind. Phrases like Eteocles and Polynices killed each other are elliptical for Eteocles and Polynices killed—each the other. Here the second proposition expands and explains the first, whilst the first supplies the verb to the second. Each, however, is elliptic. The first is without the object, the second without the verb. That the verb must be in the plural (or dual) number, that one of the nouns must be in the nominative case, and that the other must be objective, is self-evident from the structure of the sentence; such being the conditions of the expression of the idea. An aposiopesis takes place after a plural verb, and then there follows a clause wherein the verb is supplied from what went before.

When words equivalent to each other coalesce, and become compound; it is evident that the composition is of a very peculiar kind. Less, however, for these matters than for its value in elucidating the origin of certain deponent verbs does the expression of reciprocal action merit the notice of the philologist. In the latter part of the paper it will appear that for one branch of languages, at least, there is satisfactory evidence of a reflective form having become reciprocal, and of a reciprocal form having become deponent; this latter word being the term for those verbs whereof the meaning is active, and the form passive.

Beginning with those methods of denoting mutual action where the expression is the least explicit and unequivocal, it appears that in certain languages the reciprocal character of the verb is implied rather than expressed. Each man looked at his brother—or some equivalent clause, is the general phraseology of the Semitic languages.

More explicit than this is the use of a single pronoun (personal, possessive, or reflective) and of some adverb equivalent to the words mutually, interchangeably, &c. This is the habit of the Latin language,—Eteocles et Polynices invicem se trucidaverunt: also of the French, although not invariably, e. g. s'entr'aimer, s'entredire, s'entrebattre: also of the Mœso-Gothic—galeikái sind barnam tháim vôpjandam seina missô = ὁόμοιοί ἐισι παιδίοις τοῖς προσφωνοῦσιν ἁλλήλοις = loquentibus ad invicem.—Luc. vii. 32. Deutsche Grammatik, iv. 322, and iii. 13. The Welsh expressions are of this kind; the only difference being that the adverb coalesces with the verb, as an inseparable particle, and so forms a compound. These particles are dym, cym, or cy and ym. The former is compounded of dy, signifying iteration, and ym denoting mutual action; the latter is the Latin cum. Hence the reciprocal power of these particles is secondary: e. g. dymborthi, to aid mutually; dymddadlu, to dispute; dymgaru, to love one another; dymgoddi, to vex one another; dymgredu, to trust one another, or confide; dymguraw, to strike one another, or fight; çyçwennys, to desire mutually; cydadnabod, to know one another; cydaddawiad, to promise mutually; cydwystlaw, to pledge; cydymadrawn, to converse; cydymdaith, to accompany; ymadroddi, to discourse; ymaddaw, to promise; ymavael, to struggle; ymdaeru, to dispute, &c.

The form, which is at once current, full, and unequivocal, is the one that occurs in our own, and in the generality of languages. Herein there are two nouns (generally pronouns), and the construction is of the kind exhibited above—ἁλλήλους, each other, einander, l'un l'autre, &c.

Sometimes the two nouns remain separate, each preserving its independent form. This is the case in most of the languages derived from the Latin, in several of the Slavonic and Lithuanic dialects, and in (amongst others) the Old Norse, the Swedish, and the Danish,—l'un l'autre, French; uno otro, Span.; geden druheho, Bohemian; ieden drugiego, Polish; wiens wienâ, Lith.; weens ohtru, Lettish; hvert annan (masc.), hvert annat (neut.) Old Norse. See D. G. iii. 84.

Sometimes the two nouns coalesce, and form words to which it would be a mere refinement to deny the name of compounds: this is the case with the Greek—ἁλλήλον, ἁλλήλοις, ἁλλήλους.

Sometimes it is doubtful whether the phrase consist of a compound word or a pair of words. This occurs where, from the want of inflection, the form of the first word is the same in composition as it would have been out of it. Such is the case with our own language: each-other, one-another.

Throughout the mass of languages in general the details of the expression in question coincide; both subject and object are almost always expressed by pronouns, and these pronouns are much the same throughout. One, or some word equivalent, generally denotes the subject. Other, or some word equivalent, generally denotes the object, e. g. they struck one another. The varieties of expression may be collected from the following sketch:—

1. a. The subject is expressed by one, or some word equivalent, in most of the languages derived from the Latin, in several of the Slavonic dialects, in Lithuanic and Lettish, in Armenian, in German, in English, and doubtlessly in many other languages—l'un l'autre, Fr.; uno otro, Sp.; ieden drugiego, Polish; wiens wienâ, Lith.; weens ohtru, Lett.; me mæants, Armenian; einander, Germ.; one another, Engl.

b. By each, or some equivalent term, in English, Dutch, and the Scandinavian languages—each other, English; elkander, Dutch; hverandre, Icelandic, Danish, Swedish.

c. By this, or some equivalent term, in Swedish and Danish (hinanden); in Lithuanic (kitts kittâ), and in Lettish (zitts zittu).

d. By other, or some equivalent term, in Greek and Armenian; ἁλλήλους, irærats.

e. By man, used in an indefinite sense and compounded with lik in Dutch, malkander (mal-lik manlik).

f. By a term equivalent to mate or fellow in Laplandic—gòim gòimeme.—Rask, 'Lappisk Sproglære,' p. 102. Stockfleth, 'Grammatik,' p. 109.

2. a. In the expression of the object the current term is other or some equivalent word. Of this the use is even more constant than that of one expressive of the subject—l'un l'autre, French; uno otro, Spanish; ἁλλήλους, Greek; geden druheho, Bohemian; ieden drugiego, Polish; weens ohtru, Lettish; irærats, Armenian; einander, German; each other, one another, English.

b. In Lithuanic the term in use is one; as, wiens wienâ. The same is the case for a second form in the Armenian mimœan.

c. In Laplandic it is denoted in the same as the subject; as gòim gòimeme.

Undoubtedly there are other varieties of this general method of expression. Upon those already exhibited a few remarks, however, may be made.

1. In respect to languages like the French, Spanish, &c., where the two nouns, instead of coalescing, remain separate, each retaining its inflection, it is clear that they possess a greater amount of perspicuity; inasmuch as (to say nothing of the distinction of gender) the subject can be used in the singular number when the mutual action of two persons (i. e. of one upon another) is spoken of, and in the plural when we signify that of more than two; e. g. ils (i. e. A and B) se battaient—l'un l'autre: but ils (A, B, C and D,) se battaient—les uns les autres. This degree of perspicuity might be attained in English and other allied languages by reducing to practice the difference between the words each and one; in which case we might say A and B struck one another, but A, B and C struck each other. In the Scandinavian languages this distinction is real; where hinanden is equivalent to l'un l'autre, French; uno otro, Spanish: whilst hverandre expresses les uns les autres, French; unos otros, Spanish. The same is the case in the Laplandic.—See Rask's Lappisk Sproglære, p. 102.

2. An analysis of such an expression as they praise one another's (or each other's) conduct, will show the lax character of certain forms in the Swedish. Of the two pronouns it is only the latter that appears in an oblique case, and this necessarily; hence the Swedish form hvarsannars is illogical. It is precisely what one's another's would be in English, or ἄλλων ἄλλων for ἁλλήλων in Greek. The same applies to the M. H. G. einen anderen. D. G. iii. 83.

3. The term expressive of the object appears in three forms, viz. preceded by the definite article (l'un l'autre), by the indefinite article (one another), and finally, standing alone (each other, einander). Of these three forms the first is best suited for expressing the reciprocal action of two persons (one out of two struck the other); whilst the second or third is fittest for signifying the reciprocal action of more than two (one out of many struck, and was struck by, some other).

The third general method of expressing mutual or reciprocal action is by the use of some particular form of the verb. In two, and probably more, of the African languages (the Woloff and Bechuana) this takes place. In the Turkish there is also a reciprocal form: as sui-mek, to love; baki-mek, to look; sui-sh-mek, to love one another; baki-sh-mek, to look at one another; su-il-mek, to be loved; sui-sh-il-mek, to be loved mutually.—David's Turkish Grammar.

The fourth form of expression gives the fact alluded to at the beginning of the paper: viz. an instrument of criticism in investigating the origin of certain deponent verbs. In all languages there is a certain number of verbs denoting actions, reciprocal or mutual to the agents. Such are the words embrace, converse, strive against, wrestle, fight, rival, meet, and several more. There are also other words where the existence of two parties is essential to the idea conveyed, and where the notion, if not that of reciprocal action, is akin to it; viz. reproach, compromise, approach, &c. Now in certain languages (the Latin and Greek) some of these verbs have a passive form; i. e. they are deponents,—loquor, colloquor, luctor, reluctor, amplector, suavior, osculor, suspicor, Latin; φιλοτιμέομαι, φιλοφρονέομαι, μάχομαι, διαλέγομαι, ἁλέομαι, διαλύομαι, ἁμείβομαι &c., Greek. Hence arises the hypothesis, that it is to their reciprocal power on the one hand, and to the connexion between the passive, reflective and reciprocal forms on the other, that these verbs owe their deponent character. The fact essential to the probability of this hypothesis is the connexion between the reflective forms and the reciprocal ones.

Now for one branch of languages this can be shown most satisfactorily. In Icelandic the middle voice is formed from the active by the addition of the reflective pronoun, mik, me, sik, him or self. Hence it is known by the terminations mc and sc, and by certain modifications of these affixes, viz. st, s, z, mz, ms. In the oldest stage of the language the reflective power of the middle voice, to the exclusion of a passive sense, is most constant: e. g. hann var nafnadr=he had the name given him; hann nefnist=he gave as his name, or named himself. It was only when the origin of the middle form became indistinct that its sense became either passive or deponent; as it generally is in the modern tongues of Denmark, Norway and Sweden. Now in the modern Scandinavian languages we have, on the one hand, certain deponent forms expressive of reciprocal action; whilst on the other we have, even in the very earliest stages of the Old Norse, middle or reflective forms used in a reciprocal sense. Of some of these, examples will be given: but the proof of their sense being reciprocal will not be equally conclusive in all. Some may perhaps be looked on as deponents (ættust, beriast, skiliast, mödast); whilst others may be explained away by the assumption of a passive construction (fundoz=they were found, not they found each other). Whatever may be the case with the words taken from the middle and modern stages of the language, this cannot be entertained in regard to the examples drawn from the oldest Norse composition, the Edda of Sæmund. For this reason the extracts from thence are marked Edd. Sæm., and of these (and these alone) the writer has attempted to make the list exhaustive. The translations in Latin and Danish are those of the different editors.

1. Ættust, fought each other.

2. Beriaz, strike each other.

Voluspa, 41. Edd. Sæm.

This word is used in almost every page of the Sagas as a deponent signifying to fight: also in the Feroic dialect.

3. Bregþaz, interchange.

Helga-Qviþa Hundlingsbana, i. 41. ii. 26. Edd. Sæm.

4. Drepiz, kill one another.

finnuz þeir báder daudir—— en ecki vapn höfþu þeir nema bitlana af hestinum, ok þat hygia menn at þeir (Alrek and Eirek) hafi drepiz þar med. Sva segir Ðiodolfr.; "Drepaz kvádu."—Heimskringla. Ynglinga-Saga, p. 23.

The brothers were found dead—and no weapons had they except the bits of their horses, and men think they (Alrek and Eirek) had killed each other therewith. So says Thiodolf.: "They said that they killed each other."

5. Um-faþmaz, embrace each other. See Atla-Quiþa hin Grænslenzko, 42.—Edd. Sæm.

6. Földes, fell in with each other.—Om morgonet effter földes wy in Kobenhaffn.—Norwegian Letters in 1531, A. D. See Samlingar til det Norske Folks Sprog og Historie, I. 2. 70. The morning after we fell in with each other in Copenhagen.

7. Funduz, found each other, met. See Vafþrudnis-mal 17.—Sigurd-Quiþ. i. 6. Edd. Sæm.—Fareyingar-Saga, p. 44. Ðeir funduz is rendered de fandt hverandre=they found each other, in Haldorsen's Lexic. Island.

Harbards-l: 24. Edd. Sæm.

8. Gættuz, consult each other. See Voluspa, 6. 9. 21. 23. Edd. Sæm.

9. Glediaz, rejoice each other.

Rigsmal. 41.

The middle form and reciprocal sense of erost is remarkable in this passage.

10. Hauggvaz, hack each other, fight.

Vafþrudnis-Mal. 41. Edd. Sæm.

Sig-Qvið. ii. 1. Edd. Sæm.

11. Hættaz, cease.

Harbardslióð, 51. Edd. Sæm.

Such is the translation of the editors, although the reciprocal power is not unequivocal.

12. Hittaz, hit upon each other, meet. Hittoz, Voluspa, 7. Hittomk, Hadding-skata, 22. Hittaz, Solar-l: 82. Edd. Sæm. Hittust, Ol. Trygv. Sag. p. 90. Hittuz oc beriaz, Heimskringla, Saga Halfd. Svart. p. 4. Hittuz, Yngl. Sag. p. 42. alibi passim þeir hittu is rendered, in Bjorn Haldorsen's Islandic Lexicon, de traf hinanden, they hit upon each other.

13. Kiempis, fight each other,

gaar udi gaarden oc kiempis, oc nelegger hver hinanden, goes out in the house and fight each the other, and each knocks down the other.

Such is the translation by Resenius, in modern Danish, of the following extract from Snorro's Edda, p. 34.—Ganga ut i gardinn og beriast, og fellar huor annar. Here the construction is not, they fell (or knock down) each the other, but each fells the other; since fellar and nelegger are singular forms.

14. Mælast, talk to each other, converse. Talast, ditto.

Mæliz þu. Vafþrudnismal, 9.

melomc i sessi saman=colloquamur sedentes. ib. 19. Edd. Sæm.

mælast þeir vid, ádr þeir skiliast, at þeir mundi þar finnast þa,—Fóstbrædra-Saga, p. 7.

they said to each other before they parted from each other that they should meet each other there.

Yngvi ok Bera satu ok töluduz vidr.—Heimskr. Yngl. S. p. 24.

Griss mælti; hverír ero þessir menn er sva tulast vid bliðliga? Avàldi svarar; þa er Hallfreydr Ottarson ok Kolfinna dóthir min. Ol. Trygyv. Saga, p. 152. Griss said, who are these persons who talk together so blithely? Avaldi answers, they are Halfrid Ottarson and Kolfinna my daughter. Talast is similarly used in Feroic. Kvödust, bespoke each other, occurs in the same sense—þat var einn dag at Brand ok Finbogi fundust ok kvödust blídliga.—Vatnsdæla-Sag. p. 16.

15. Mettæst, meet each other, meet.

Kungen aff Ffranchriche, kungen aff England, oc kungen aff Schottland skule motes til Chalis.—Letter from Bergen in 1531, from Samlinger til det Norske Folks Sprog og Historie, i. 2. p. 53. The king of France, the king of England, and the king of Scotland should meet each other at Calais.

Throughout the Danish, Swedish and Feroic, this verb is used as a deponent.

16. Rekaz, vex each other.

Rigsmal. 32. Edd. Sæm.

17. Sakaz, accuse each other, recriminate.

Hamdis-Mal. 28.

Ægis-drecka, 5.

Ibid. 19. Edd. Sæm.

18. Saz, looked at each other.

Rigsmal. 24.

19. Sættaz, settle between each other, reconcile.—Atla-Mal. 45. Edd. Sæm.

Komu vinir þveggia þvi vid, at þeir sættuz, ok lögdu konungar stefnu med sér, ok hittuz ok gérdo frit mellum sin.—Heimsk. Yngling-S. 42.

There came friends of both in order that they should be reconciled, and the kings sent messages between them, and met and made peace between them.—Also Vatnsd. S. p. 16.

20. Seljas, to give to each other.

Sig. Qv. iii. 1. Edd. Sæm.

21. Sendaz, send, or let pass between each other.

Atla-Mal. 85.

Here, over and above the use of senduz and henduz, ser is equivalent to hinanden.

22. Skiliaz, part from each other.

Skiliumz.

Solar-Lioð. 82.

Skiliaz.

Sigurd-Qviþ. i. 24.

Skiliomc.

Ibid. 53. Edd. Sæm.

Vit sjiljiast, we two part—

Occurs in the poem Brinilda (st. 109) in the Feroic dialect. In Danish and Swedish the word is deponent.

23. Skiptust, interchange.

Ðeir skiptust mörgum giöfum vid um vetrinn—Vatns-dæla-S. 10. they made interchanges with each other with many gifts for the winter.

Also in the Feroic.

24. Strujast, strike one another, fight. Feroic.

og mötast tair, og strujast avlaji lanji.—Fareying-Sag. 18. Feroic text.

ok mætast þeir, ok berjast mjök leingi.—Icelandish text.

de mödtes og strede meget længe imod hinanden.—Danish text.

they met and fought long against each other.

at e vilde vid gjordust stålbröir, og strujast ikkji longur.—Feroic text, p. 21.

at við gerðimst fèlagar, en berjumst eigi leingr.—Icelandic text.

at vi skulle blive Stalbröde og ikke slaaes længer—Danish text.

that we should become comrades and not fight longer.

The active form occurs in the same dialect:

tajr struija nú langji.

18.

25. Truasc, trust each other.

vel mættern þæir truazc.

För Skirnis. Edd. Sæm.

26. Unnaz. See Veittaz.

27. Vegiz, attack each other.

vilcat ec at iþ reiþir vegiz.

Ægisdrecka 18. Edd. Sæm.

I will not that ye two angry attack each other.

28. Veittaz, contract mutually.

þav Helgi ok Svava veittuz varar, ok unnoz forþo mikit=Helgius et Svava pactum sponsalitium inter se contraxerunt, et alter alterum mirifice amarunt.—Haddingia-Sk. between 29 and 30.

29. Verpaz, throw between each other.

urpuz á orþom.

Atl.-M. 39. Edd. Sæm.

verba inter se jaciebant.

Such is a portion of the examples that prove the reciprocal power of the reflective or middle verb in the language of Scandinavia; and that, during all its stages and in each of its derived dialects. It cannot be doubted that to this circumstance certain verbs in Danish and Swedish owe their deponent form: viz. vi slåss, we fight (strike one another); vi brottas, we wrestle; vi omgass, we have intercourse with; vi mötas, we meet, Swedish; vi slaaes, we fight; vi skilles, we part; vi mödes, we meet, Danish. In the latest Swedish grammar, by C. L. Daae, this reciprocal (vekselvirkende) power is recognized and exhibited. See Udsigt over det Svenske Sprogs Grammatik. Christiana, 1837. The same is the Molbech's Danske Ordbog in vv. skilles, slaaes, mödes.

Next to the Norse languages the French affords the best instances of the reciprocal power of the reflective verb; as se battre, s'aimer, s'entendre, se quéreller, se reconcilier, se disputer, and other words of less frequent occurrence.

Ces enfans s'aimaient, s'adoraient, se sont jetés à mes pieds en pleurant.—Les Inséparables, A. 1. S. 1.

Les Républics Italiens acharnés à se détruire.—Pardessus II. 65.

This has been recognized by an old grammarian, Restaut, who insists upon the use of the adverb entre, in order to avoid the ambiguity of such phrases as "vous vous dites des injures;" "nous nous écrivons souvent;" "Pierre et Antoine se louent à tout moment."

By a writer in the Museum Criticum the reciprocal power of the Greek middle has been indicated. For the classical languages the question has not met with the proper investigation. Passages where the sense is at least as reciprocal as in the line

Χεῖρος τ' ἁλλήλων λαβήτην καὶ μιστώσαντο.—Il. vi. 233,
must be numerous.

In the Dutch language the use of zich for elkander is a peculiarity of the Guelderland and Overyssel dialects; as "zij hebt zich eslagen," for "zij hebben elkander geslagen." See Opmerkingen omtrent den Gelderschen Tongval, in Taalkundig Magazijn ii. 14. p. 403.

Of the use of ser for hinanden or hverandre, when uncombined with the verb, we have, amongst other, the following example in the Icelandic version of the Paradise Lost:—

B. 6.

Similar to this are the phrases vi se os igjen, we see us (each other) again, in Danish, and wir sehen uns wieder, in German. Examples from the M. H. G. are given in the D. G. iv. The Turkish sign of the reciprocal verb is identical with the demonstrative pronoun, i. e. ش. This may possibly indicate a connection between the two forms.

Other points upon the subject in hand may be collected from the Deutsche Grammatik, iii. 13. 82; iv. 454. Here the adverbial character of the M. H. G. einander for einandern, the omission of ein, as in anander for an einander, and the omission (real or supposed) of ander in "wider ein=wider einander," are measures of the laxity of language caused by the peculiarity of the combination in question. At present it is sufficient to repeat the statement, that for one group of languages at least there is satisfactory proof of certain deponents having originally been reciprocal, and of certain reciprocal expressions having originally been reflective.

ON THE CONNEXION BETWEEN THE IDEAS OF ASSOCIATION AND PLURALITY AS AN INFLUENCE IN THE EVOLUTION OF INFLECTION.

READ
BEFORE THE PHILOLOGICAL SOCIETY.
MARCH 9, 1849.

It is well-known that by referring to that part of the Deutsche Grammatik which explains those participial forms which (like y-cleped in English, and like ge-sprochen and the participles in general in German) begin with ge or y, the following doctrines respecting this same prefix may be collected:—

1. That it has certainly grown out of the fuller forms ka or ga.

2. That it has, probably, grown out of a still fuller form kam or gam.

3. That this fuller form is the Gothic equivalent of the Latin cum=with.

Such are the views respecting the form of the word in question. Respecting its meaning, the following points seem to be made out:—

1. That when prefixed to nouns (as is, not rarely, the case), it carries with it the idea of association or collection:—M. G. sinþs=a journey, ga-sinþa=a companion; O. M. G. perc=hill; ki-pirki=(ge-birge) a range of hills.

2. That it has also a frequentative power. Things which recur frequently recur with a tendency to collection or association:—M. H. G. ge-rassel=rustling; ge-rumpel=crumpling.

3. That it has also the power of expressing the possession of a quality:—

This is because every object is associated with the object that possesses it—a sea with waves=a wavy sea.

The present writer has little doubt that the Tumali grammar of Dr. Tutshek supplies a similar (and at the same time a very intelligible) application of a particle equivalent to the Latin cum.

He believes that the Tumali word=with is what would commonly be called the sign of the plural number of the personal pronouns; just as me-cum and te-cum would become equivalents to nos and vos, if the first syllables were nominative instead of oblique, and if the preposition denoted indefinite conjunction. In such a case

Such is the illustration of the possible power of a possible combination. The reasons for thinking it to have a reality in one language at least lie in the following forms:—

1. The Tumali word for with is da.

2. The Tumali words for I, thou, and he respectively are ngi, ngo, ngu.

3. The Tumali words for we, ye, they are ngin-de, ngon-da, ngen-da respectively.

4. The Tumali substantives have no such plural. With them it is formed on a totally different principle.

5. The Tumali adjectives have no plural at all.

6. The Tumali numerals (even those which express more than unity and are, therefore, naturally plural) have a plural. When, however, it occurs, it is formed on the same principle as that of the plurals of the substantive.

7. The word da=with is, in Tumali, of a more varied application than any other particle; and that both as a pre-position and a post-position:—daura=soon (da=in, aura=neighbourhood); datom=in (with) front (face); d-ondul=roundabout (ondul=circle); dale=near (le=side), &c.

8. Prepositions, which there is every reason to believe are already compounded with da, allow even a second da, to precede the word which they govern:—daber deling=over the earth (ber=earth).

9. The ideas with me, with thee, with him, are expressed by ngi-dan, ngo-dan, and ngu-dan respectively; but the ideas of with us, with you, with them are not expressed by nginde-dan, ngonda-dan, ngenda-dan; but by peculiar words—tinem=with us; toman=with you; tenan=with them.

On the other hand, the following fact is, as far as it goes, against this view, a fact upon which others may lay more stress than the present writer. "Da admits of a very varied application. Respecting its form the following should be observed: (a.) That a may be elided when it happens to stand as a preposition before words which begin with a vowel: for instance, ardgen, 'the valley'; dardgen, 'in the valley'; ondul, 'the circle'; dondul, 'round about in the circle'. (b.) It changes its a into ê, e, i, o, u, according to the vowel of the syllable before which the da is placed, or even without any regard to it. Instances of this are found in diring, dorong, &c.; further instances are, doromko, 'into the hut' (rom); dètum or dotum, 'in the grave.' (c.) As a postposition it appends an n: adgdan, 'on the head'; aneredan, 'on the day.'" Taking the third of these rules literally, the plural pronouns should end in dan rather than in da and de.

It is considered that over and above the light that this particular formation (if real) may throw upon the various methods by which an inflection like that of the plural number may be evolved, and more especially upon the important but neglected phænomena of the so-called inclusive and exclusive plurals, many other points of general grammar may be illustrated.

ON THE WORD CUJUM.

READ
BEFORE THE PHILOLOGICAL SOCIETY,
MARCH 9, 1849.

The writer wishes to make the word cujum, as found in a well-known quotation from the third eclogue of Virgil,—

Dic mihi Damæta cujum pecus?

the basis of some remarks which are meant to be suggestions rather than doctrines.

In the second edition of a work upon the English language, he devoted an additional chapter to the consideration of the grammatical position of the words mine and thine, respecting which he then considered (and still considers) himself correct in assuming that the current doctrine concerning them was, that they were, in origin, genitive or possessive cases, and that they were adjectives only in a secondary sense. Now whatever was then written upon this subject was written with the view of recording an opinion in favour of exactly the opposite doctrine, viz. that they were originally adjectives, but that afterwards they took the appearance of oblique cases. Hence for words like mine and thine there are two views:—

1. That they were originally cases, and adjectives only in a secondary manner.

2. That they were originally adjectives, and cases only in a secondary manner.

In which predicament is the word cujum? If in the first, it supplies a remarkable instance of an unequivocally adjectival form, as tested by an inflection in the way of gender, having grown out of a case. If in the second, it shows how truly the converse may take place, since it cannot be doubted that whatever in this respect can be predicated of cujus can be predicated of ejus and hujus as well.

Assuming this last position, it follows that if cujus be originally a case, we have a proof how thoroughly it may take a gender; whereas if it be originally an adjective, ejus and hujus (for by a previous assumption they are in the same category) are samples of the extent to which words like it may lose one.

Now the termination -us is the termination of an adjective, and is not the termination of a genitive case; a fact that fixes the onus probandi with those who insist upon the genitival character of the words in question. But as it is not likely that every one lays so much value upon this argument as is laid by the present writer, it is necessary to refer to two facts taken from the Greek:—

1. That the class of words itself is not a class which (as is often the case) naturally leads us to expect a variation from the usual inflections. The forms οὗ, οἷ, ἕ, and ὅς, οὗ, ὧ, are perfectly usual.

2. That the adjectives ὃς = ἑὸς,[1] κοῖος = ποῖος, and ὁῖος, are not only real forms, but forms of a common kind. Hence, if we consider the termination -jus as a case-ending, we have a phænomenon in Latin for which we miss a Greek equivalent; whilst on the other hand, if we do not consider it as adjectival, we have the Greek forms ὁῖος, κοῖος = ποῖος and ὃς = ἑὸς, without any Latin ones. I do not say that this argument is, when taken alone, of any great weight. In doubtful cases, however, it is of value. In the present case it enables us to get rid of an inexplicable genitival form, at the expense of a slight deflection from the usual power of an adjective. And here it should be remembered that many of the arguments in favour of a case becoming an adjective are (to a certain extent) in favour of an adjective becoming a case—to a certain extent and to a certain extent only, because a change in one direction by no means necessarily implies a change in the reverse one, although it is something in favour of its probability.

Probably unius, ullius, illius, and alterius, are equally, as respects their origin, adjectival forms with ejus, cujus, and hujus.

Now it must not be concealed that one of the arguments which apply to words like mine and thine being adjectives rather than genitives, does not apply to words like ejus, cujus, and hujus. The reason is as follows; and it is exhibited in nearly the same words which have been used in the work already mentioned.—The idea of partition is one of the ideas expressed by the genitive case. The necessity for expressing this idea is an element in the necessity for evolving a genitive case. With personal pronouns of the singular number the idea of partition is of less frequent occurrence than with most other words, since a personal pronoun of the singular number is the name of a unity, and, as such, the name of an object far less likely to be separated into parts than the name of a collection. Phrases like some of them, one of you, many of us, any of them, few of us, &c., have no analogues in the singular number, such as one of me, a few of thee, &c. The partitive words that can combine with singular pronouns are comparatively few, viz. half, quarter, part, &c.; and they can all combine equally with plurals—half of us, a quarter of them, a portion of us. The partition of a singular object with a pronominal name is of rare occurrence in language. This last statement proves something more than appears at first sight. It proves that no argument in favour of the so-called singular genitives, like mine and thine, can be drawn from the admission (if made) of the existence of the true plural genitives ou-r, you-r, the-ir. The two ideas are not in the same predicament.

Again, the convenience of expressing the difference between suus and ejus, is, to a certain extent, a reason for the evolution of a genitive case to words like is; but it is a reason to a certain extent only, and that extent a small one, since an equally convenient method of expressing the difference is to be found in the fact of there being two roots for the pronouns in question, the root from which we get ea, id, eum, ejus, &c., and the root from which we get sui, sibi, suus, &c.

Here the paper should end, for here ends the particular suggestion supplied by the word in question. Two questions however present themselves too forcibly to be wholly passed over:—

I. The great extent to which those who look in Latin for the same inflections that occur in Greek, must look for them under new names. That two tenses in Greek (the aorist like ἔ-τυπ-σα, and the perfect like τέ-τυφ-α) must be looked for in the so-called double form of a single tense in Latin (vic-si, mo-mordi) is one of the oldest facts of this sort. That the Greek participle in -μενος (τυπτόμενος) must be sought for in the passive persons in -mini is a newer notice.

II. The fact that the character of the deflection that takes place between case and adjective is not single but double. It goes both ways. The change from case to adjective is one process in philology; the change from adjective to case another; and both should be recognized. This is mentioned for the sake of stating, that except in a few details, there is nothing in the present remarks that is meant to be at variance with the facts and arguments of five papers already laid before this Society, viz. those of Mr. Garnett on the Formation of Words from Inflected Cases, and on the Analysis of the Verb.

The papers alluded to really deal with two series of facts:—(A.) Deflection with identity of form.—In this the inflection is still considered an inflection, but is dealt with as one different from what it really is, i. e. as a nominative instead of an oblique one. Some years back the structure of the Finlandic suggested to the present writer:—

1. A series of changes in meaning whereby such a term as with waves might equal wavy.

2. The existence of a class of words of which sestertium was the type, where an oblique case, with a convertible termination, becomes a nominative.

3. The possible evolution of forms like fluctuba, fluctubum=fluctuosa, fluctuosum, from forms like fluctubus.

Mr. Garnett has multiplied cases of this kind; his illustrations from the Basque being pre-eminently typical, i. e. like the form sestertium. If the modern vehicle called an omnibus had been invented in ancient Rome, if it had had the same name as it has now, and if its plural form had been omnibi, it would also have been a typical instance.

Words of the hypothetical form fluctuba, fluctubum, have not been discovered. They would have existed if the word just quoted had been (if used in ancient Rome at all) used as an adjective, omnibus currus, omniba esseda, omnibum plaustrum.

(B.) Deflection with superaddition.—Here the inflection is dealt with as if it were not inflectional but radical. This is the case with ἴφιος. Words like it-, as proved by the genitive i-t-s, and the so-called petrified (versteinerte) nominative cases of the German grammarians, are of this class.

ON THE AORISTS IN -KA.

READ
BEFORE THE PHILOLOGICAL SOCIETY,
MARCH 11, 1853.

A well-known rule in the Eton Greek Grammar may serve to introduce the subject of the present remarks:—"Quinque sunt aoristi primi qui futuri primi characteristicam non assumunt: ἔθηκα posui, ἔδωκα dedi, ἥκα misi, εἴπα dixi, ἥνεγκα tuli." The absolute accuracy of this sentence is no part of our considerations: it has merely been quoted for the sake of illustration.

What is the import of this abnormal κ? or, changing the expression, what is the explanation of the aorist in -κα? Is it certain that it is an aorist? or, granting this, is it certain that its relations to the future are exceptional?

The present writer was at one time inclined to the doubts implied by the first of these alternatives, and gave some reasons[2] for making the form a perfect rather than an aorist. He finds, however, that this is only shifting the difficulty. How do perfects come to end in -κα? The typical and unequivocal perfects are formed by a reduplication at the beginning, and a modification of the final radical consonant at the end of words, τύπ(τ)ω, τέ-τυφ-α; and this is the origin of the χ in λέλεχα, &c., which represents the γ of the root. Hence, even if we allow ourselves to put the κ in ἔθηκα in the same category with the κ in ὀμώμοκα, &c., we are as far as ever from the true origin of the form.

In this same category, however, the two words—and the classes they represent—can be placed, notwithstanding some small difficulties of detail. At any rate, it is easier to refer ὀμώμοκα and ἔθηκα to the same tense than it is to do so with ὀμώμοκα and τέτυφα.

The next step is to be sought in Bopp's Comparative Grammar. Here we find the following extract:—"The old Slavonic dakh 'I gave,' and analogous formations remind us, through their guttural, which takes the place of a sibilant, of the Greek aorists ἔθηκα, ἔδωκα, ἧκα. That which in the old Slavonic has become a rule in the first person of the three numbers, viz. the gutturalization of an original s, may have occasionally taken place in the Greek, but carried throughout all numbers. No conjecture lies closer at hand than that of regarding ἔδωκα as a corruption of ἔδωσα," &c.... "The Lithuanian also presents a form which is akin to the Greek and Sanscrit aorist, in which, as it appears to me, k assumes the place of an original s." (vol. ii. p. 791, Eastwick's and Wilson's translation.) The italics indicate the words that most demand attention.

The old Slavonic inflection alluded to is as follows:—