Plat. vol. Iv. p. 70. ὁ μὴ ἔχων ΚΑΚΙΑΝ. Plat. vol. iv. p. 70. ὁ ἔχων ΑΔΙΚΙΑΝ. Ibid. p. 57. ἆρ ̓ ἂν τυγχάνη ΔΙΚΗΣ τε καὶ ΤΙΜΩ ΡΙΑΣ;

Demosth. vol. I. p. 142. ταῦτ ̓ ΑΠΙΣΤΙΑΝ, ταῦτ ̓ ΟΡΓΗΝ ἔχει.

Demosth. vol. II. p. 1232. ἂν δὲ ληφθῶσι, ΣΥΓ ΓΝΩΜΗΣ τυχεῖν.

Plut. Conviv. p. 93. ανεπλήσθη τὸ πρόσωπον ΕΡΥΘΗΜΑΤΟΣ.

Demosth. vol. I. p. 151. ΚΟΛΑΚΕΙΑΣ καὶ ΒΛΑΒΗΣ καὶ ΑΠΑΤΗΣ λόγος μεστός.

The same usage prevails where the Nouns are names of substances.

Verbs of partaking do also for the most part, though not invariably, follow the same rule: the reason of the uncertainty seems to be, that usually they are employed merely in the sense of having, though if they were used strictly in the sense of having or dividing with others, the Abstract Nouns subjoined to them might take the Article; for though attributes and qualities are wholes, which no single individual can claim to the exclusion of every other, yet of these wholes he may be a partaker, and in truth is so of every attribute, which can be ascribed to him even in the smallest degree: however, it was to be expected, for the reason alleged, that the anarthrous use would be by far the more

common.

On the same principle it is, that in the common phrases ἄνοιαν, αἰσχύνην, &c. όφλισκάνειν, δίκην διδόναι, novxíav ayew, and many more, the Article is invari

ably omitted'. Since in many of these phrases two words are employed to convey the meaning of one, and in all of them a single Verb may be imagined, which would express the meaning, I shall consider this as a Hendiadys, and shall hereafter refer to what is here said of all such phrases, under that appellation.

§ 2. In the same manner we may account for the anarthrous use of Abstract Nouns, when they are employed in the Dative Case adverbially. In this sense they are of very common occurrence, and are sometimes so joined with real Adverbs, that their import cannot be mistaken: thus in the first and fourth of the following

EXAMPLES.

Eurip. Orest. p. 191. ΔΙΚΑι μέν, καλῶς δ ̓ οὔ Arist. de Mor. Nic. lib. vI. c. 3. 'YпOAHYEI Kai ΔΟΞΗ, ἐνδέχεται διαψεύδεσθαι.

Demosth. vol. I. p. 41. ΦΥΣΕΙ δ υπάρχει τοῖς παροῦσι τὰ τῶν ἀπόντων.

Thucyd. lib. v. § 70. ἐντόνως καὶ ΟΡΓΗι χωροῦντες. Plato, vol. iv. p. 89. οὔτε ΣΟΦΙΑΣ ΕΝΔΕΙΑ. οὔτ ̓ ΑΙΣΧΥΝΗΣ ΠΕΡΙΟΥΣΙΑ..

In these Examples it is to be observed, that the manner, in which any thing is said to happen or be done, is not spoken of with reference to any particular subject, to which such manner is more espe

1 Yet we find ΤΗΝ εἰρήνην, ΤΗΝ σύμβασιν, ΤΑΣ ἀνοχὰς Toiciodai. In such phrases, however, there is, probably, a reference to the war, the termination or suspension of which is in question.

cially attributable. But the case may be otherwise : the manner may be adverted to as being the attribute more especially of the subject in question: and then the Article will be prefixed, and will, as in the instances already mentioned, have the force of a Possessive Pronoun.

EXAMPLES.

Arist. Rhet. lib. II. cap. 15. (wσ THI μvýun μᾶλλον ἢ ΤΗι ἐλπίδι.

Thucyd. lib. v. § 72. ΤΗι ἐμπειρίᾳ Λακεδαιμόνιοι ἐλασσωθέντες τότε, ΤΗι ἀνδρείᾳ ἔδειξαν οὐχ ἧσσον περιγενόμενοι 2.

On the whole, it appears that Abstract Nouns for the most part refuse the Article, never taking it, excepting in the four cases before exemplified. The only caution requisite respects the more or less abstract sense, in which these Nouns may be used. Many instances will occur, in which they are anarthrous, where, had they been used in the more abstract sense, the proposition would still have been true. Such passages are not to be subjected to the rashness of conjectural emendation. It was sufficient for the writer, if his assertion were likely

? In this passage, it may be supposed, that both eμreipia and avopeia should, according to what has been advanced above, be anarthrous. Baver, however, in his excellent edition of Thucydides, Lips. 1790. has shewn, that Tureipia must be ren

dered per artem HOSTIUM: and by τῇ ἀνδρεία we must

plainly understand "by the bravery of the Spartans." The Articles, therefore, are necessary, the Nouns not being employed in the adverbial sense, but with reference to particular subjects.

to gain assent in its limited form; and it was better to affirm in part without the danger of contradiction, even where the proposition might have been couched in the most general and unlimited terms, than to risk an extreme latitude of assertion, where it was not needed. This remark may contribute to account for the frequent absence of the Article, where unquestionably it might have been employed by the first of the four canons.

CHAP. VI.

ANOMALIES.

IT has thus far been my endeavour to investigate the nature of the Article, and to shew that its principal insertions and omissions before the several classes of Nouns are explicable on the proposed hypothesis. It was not, however, to be expected, in a case of this sort, that we should meet with no anomalies; and it will not be deemed injurious to that hypothesis, if certain usages occasionally prevail, of which it pretends not to assign the cause. It is sufficient, if they furnish no evidence of its futility: and it is to be observed, that they are omissions of the Article where it might have been inserted, not insertions irreconcileable with its alleged nature.

§ 1. It has been shewn that the Article is commonly prefixed to Nouns, which are employed кar' etoxýv, and in some similar cases noticed above: but I am not aware that any philologist has remarked how frequently such Nouns become anarthrous after Prepositions.

EXAMPLES.

Plat. Theæt. sub init. Kaтà ПОAIN, the city (Athens.)

Ibid. KaT' ATOPAN, the Forum.

I