this ambiguity can never arise in Latin, which has not only the means of distinguishing the three relations of personality or successive proximity by the indicative pronouns hic, iste, ille (Lat. Gr., 62), but can also make a reference still more perspicuous by means of the distinctive pronoun is, and its derivatives ipse and idem (Lat. Gr., 63). Above all, in the oblique narration, the use of the reflexive or reciprocal se and its possessive suus takes away all chance of a confusion of persons. In English we might find such a sentence as this:'Mr. A. has received Mr. B's letter about Mr. C's affair. He is sorry to say that he will not be able to meet him in the city to day, &c,' where it is only a knowledge of the circumstances which can enable us to understand whether A., B., or C., are meant by the 'he' and 'him.' The opposition between se, is and ille, in Latin, allows all this to be manipulated by a very simple machinery. Compare such passages as the following:-'Non est igitur amici, talem esse in eum (amicum), qualis ille in se est,' 'to act towards him (the friend, as the person last mentioned), just as the other (the friend, as more distant) is towards himself' (the subject of the sentence) (Cic. Læl., 16, 59); 'Antonius-ei (Attico)-sua manu scripsit, ne timeret, statimque ad se veniret; se eum et illius causâ Gellium Canum de proscriptorum numero exemisse' (C. Nepos., x., 4); ad eum filiam ejus adduxit, ut ille insperato adspectu, si non omnem, at aliquam partem mororis sui deponeret' (Cic. Sest., 3, 7), where the father is opposed to the daughter; nam ego, quæ sunt oratori cum illis communia, non mutuor ab illis (the philosophers); isti (those whom you are speaking of), quæ de his rebus disputant, oratorum esse concedunt' (Orat. I., 13, 55).

The relative in Latin differs from the English relative, not only in the greater flexibility and distinctness which it derives from its different cases and genders, but also from its constant use at the beginning of a distinct sentence to serve as a word of connexion; and in this case we may often find two relative

words together. As we should use the demonstrative in English, we must be careful conversely to employ the relative in Latin, however much it may seem inconsistent with our vernacular idiom. There are frequent examples of this in the exercises, and the student's attention is often called to the discrepancy in the two modes of expression.

Among the pronominal distinctions, which make the machinery of the Latin language more effective, there is none more remarkable than those of the indefinite pronouns. The English language uses the words 'some' 'any' 'every,' in a very vague and inconsistent manner; but the Latin is most precise in distinguishing between the different grades of comprehension and exclusion indicated by these words. Thus we pass regularly from quisque, 'every one,' which includes all objects referred to, to quilibet, 'any one,' which allows indefinite selection; from this to aliquis, 'some one,' which presumes a special selection, and from this to quisquam or ullus, any at all,' which allows no selection, but rather excludes from the range of choice (Lat. Gr., 68). The distinction between indefinite and special selection in quilibet and aliquis is well given in a line of Afranius (ap. Cic. Tusc. Disp., iv., 25 § 55):

Dummodo doleat aliquid, doleat quidlibet

'provided only he suffers some pain, it matters not what pain he suffers,'—i. e., ' if you will only make some special selection, take the range as wide as you please.' With regard to quisque, it must be observed that, although all are included, there is always an implication of separate individuality. As Madvig says (ad Cic. de Fin., I., 4, 11):—'Quisque semper cum aliquâ distributione singulos separatim significat.' Hence we have passages like the following (Liv., xxxviii., 23):—'Laudati quoque pro concione omnes sunt, donatique pro merito quisque,' i. e., 'they were all praised collectively, but they were rewarded according to their respective merits.' Nägelsbach, who has written sensibly on this subject (Stilistik, p. 249 sqq.), says

that the complete use of quisque may be derived from the following four examples :

1. Non omnia omnibus tribuenda sunt, sed suum cuique; 2. Omnes idem faciunt, sed optimus quisque optime; 3. Non omnibus annis hoc fit, sed tertio quoque anno; 4. Non omnes idem faciunt, sed quod quisque vult.

With respect to the last, he observes that its application is very often disregarded by modern Latinists, and it is forgotten that although, in modern style, 'every' belongs to the antecedent, it is invariably placed in the relative sentence in Latin. Thus, we should say in English 'every one in a theatre is entitled to the place he has taken; but the Latin idiom requires that we should say: 'the place, which every one has occupied, is his;' so Cic. Fin. iii. 20, 67; theatrum ut commune sit, recte tamen dici potest, ejus esse eum locum, quem quisque occupârit.' Hence in parallel clauses, we have quisque in each; as: 'quod cuique obtigit, id quisque teneat.' (Cic. Off., i. 7, 4). As a matter of style, the opposition between quisque and quispiam must be carefully observed, especially in frequentative sentences. Compare, for example, the following passages of Cæsar (B. G., v. 34):—'quoties quæque cohors procurreret, ab eâ parte magnus hostium numerus cadebat;' and (B. G. 1. 1. 35): 'quum quæpiam cohors ex orbe excesserat atque impetum fecerat, hostes velocissime fugiebant.' A. Grotefend, who has referred to this last passage (Materials, Excursus ix., ed. Arnold), shows, as usual, the imperfection of his Latin scholarship. He says, that in this example 'one could not substitute quædam cohors for quapiam cohors,' for that 'quædam cohors would convey the notion of only one cohort, thought of by Cæsar, but without further designating the particular cohort; whereas quapiam cohors conveys the notion of several, advancing at different times.' This interpretation would be justifiable, if quapiam in the one passage of Cæsar were

equivalent to quæque in the other. But the true Latin scholar cannot fail to see that, in the former passage,' quoties quæque cohors procurrerat' means as often as each cohort respectively had advanced;' but quum quæpiam cohors excesserat' signifies when some one cohort had sallied forth.' On the other hand, quidam always denotes an object known, but not named; like the 'Accurrit quidam notus mihi nomine tantum,' of Horace. Aliquis is a degree more definite than quispiam; a degree less definite than quidam. For quispiam is merely 'some one;' aliquis is 'some one in particular;' and quidam is 'some one known;' 'some one certain or fixed.' Hence quidam is used with strong epithets to indicate that the thing was really, certainly, and absolutely so; or to imply that it was so in some sort, or to a certain extent; as 'eximia quædam magnitudo,' 'a really eminent greatness;' i. e., something certain in that way, though perhaps indescribable.* In Grotefend's Excursus, already referred to, there is a most confused account of aliquis. He says: 'when the ali is quite unemphatic, aliquis is identical in meaning with quispiam. Quite different, however, is the import of aliquis, when the ali is emphatic;' and in spite of this assertion of emphasis, without which ali-quis would not differ from quis, he translates aliquis by the English 'any' in a number of passages where it follows si, and where 'any' could not be used except in an exclusive sense, like quisquam. The fact is, that the English 'some' must always be used in translating aliquis and the other words with the prefix ali-. Thus: 'si aliquid dandum est voluptati' means if some concession must be made to pleasure;' ́si aliquid de summâ gravitate Pompeius, multum de cupiditate Cæsar dimisisset,' 'if Pompey had given up somewhat of his excessive stateliness, and Cæsar had relinquished a good deal of

iv.

* On the similar use of rs in Greek, see 'Ast ad Plat. Remp., P. 491.

his eager ambition ;'
magistratu affectus est, eâ me consulem affecit,'

quâ injuriâ nemo unquam in aliquo

he offered

to me, when consul, an insult, which no one in something of an official position had ever received.' If we rendered this last passage, with Grotefend or his translator, by inserting any, we must at the same time substitute ullo for aliquo in the Latin text. Hence he is quite wrong in translating Cic., Catil. i. 8, 20: abire in aliquas terras,' by 'into any land whatever, of whatever nature it may be;' for the context no less than the meaning of aliquis shows that some place of refuge is intended.

(b) Particles.

THE general distinctions of the particles have been fully given in the Complete Latin Grammar, pp. 87-102. A practical familiarity with this part of the machinery is best acquired by practice and an assiduous perusal of the best writers; and the student will find an immense storehouse of facts in Hand's Tursellinus, which, if not always scientifically arranged, furnishes ample materials for forming a correct judgment. In fact, I am not aware that any Latin particle has given rise to much doubt or difficulty except immo or imo, which some consider as a negative, others as an affirmative adverb. 'Nulla esse videtur,' says Hand (iii. p. 218), ‘Latinæ linguæ particula, in quâ explicandâ aut usurpandâ homines nostri magis opinione falluntur, quam immo. Nam quum multi originem frustra quæsiverunt, tum significationem constituerunt ab usu antiquo alienam.' There is, however, really no excuse for this confusion of ideas. It is perfectly obvious that immo is an adverb analogous to oppido, certo, &c., from the superlative immus or imus, the double m being justified by the correlative summus, and the single m representing the older practice as seen in olim for ollim (Varronianus, p. 314, 2nd ed.). It is a matter of little consequence, whether we distinguish between the origin of infimus and imus, as