^{page 231 note 1} Heibges, S., De clausulis Charitoneis, Diss. Münster, 1911.Google Scholar

^{page 231 note 2} It may be that literature about Plutarch has escaped my notice. I hope, however, that this omission has not rendered the following article superflnous, even if it be from a methodological point of view.

^{page 231 note 3} These 128 possible cases are specified in Table I.

^{page 231 note 4} This article is meant to be preliminary, not exhaustive. Whether I am justified in taking it as anceps should be shown by a special investigation.

^{page 231 note 5} Ed. Hude (what edition is used is practically immaterial).

^{page 232 note 1} Ed. Sintenis.

^{page 232 note 2} When the number in Plutarch was smaller than in Thucydides, I put a - sign before it; in other cases a +.

^{page 232 note 3} The Moralist perhaps may prove to be not rhythmically composed.

^{page 233 note 1} I hardly feel justified in using a stronger expression here. In Cicero and other writers the case is quite different, there being more material at one's disposition.

^{page 233 note 2} Die antike Kunstprosa, p. 914: ‘² Der rhythmisch mit dem Ditrochaus identische Dispondeus,’ etc.

^{page 233 note 3} On this point sufficient stress has been laid by Bornecque {Museum XIII. 6, col. 210), although in his own works he has evidently not succeeded in avoiding the same error.

^{page 233 note 4} Of course, as to the method, and not ‘pour certain nombre de faits.’

^{page 234 note 1} For Norden see e.g. II. p. 914: ‘… So zeigt doch eine genaue Analyse des Einzelnen, dass er (Demosthenes) gewisse Rhythmenge, schlechter in den Klauseln bevorzugt. Es finden sich in der genannten Rede an den Schlüssen der Kola:

‘1. Der Ditrochaus — ⌣ — 48 mal.

‘2. Der rhythmisch mit dem Ditrochaus identische Dispondeus — — — 59 mal. (identische! ? Cf. Table VI., where it is evident that Plutarch consistency avoids this form.)

‘3. …’and so on.

It is quite evident that the shorter forms (four syllables) occur more frequently than the longer (e.g. — ⌣ ⌣ — — ⌣ ⌣ —). Norden seems to take this for an argument in favour of a preference for these forms (‘Bevorzugung’).

Attention is also drawn to p. 923 near the bottom, where the form — ⌣ ⌣ — — ⌣ ⌣ is considered as a selected or preferred clausula without any proof. (Cf. also p. 924.)

A final remark: If it should be possible to prove that not a single form of clausula has a preference in Demosthenes, would Norden then still be ready to uphold his ‘Analyse’ (p. 911, sqq.)?

^{page 234 note 2} The figures of Chariton are taken from Heibges.

^{page 235 note 1} Diss. Münster, 1910: ‘De numeri oratorii primordiis.’

^{page 236 note 1} For illustration: Kroll, B.Ph. W. 1905, p. 1659; Ed. de Jonge, Les clausuks dans Saint Cyprien, Lut. Par. 1905; Tolkiehn, B.Ph.W. 1906, p. 907; Jordan, Rhythm. Prosa, Leipz. 1905; Heibges, Diss. Münster, 1911, p. 21.

^{page 236 note 2} To understand this clearly, one should take into consideration that a form of four syllables (e.g. — ⌣ — ⌣) comprises sixteen forms of eight syllables, and, to use an a priori argumentation, has therefore sixteen times as great a chance to be found as e.g. the form — ⌣ — ⌣ ⌣ ⌣ — ⌣

^{page 236 note 3} φερ⋯ συνθ. όνομ 115.

^{page 236 note 4} The same is true for Chariton. Cf. p. 234.

^{page 237 note 1} Also the resolution — ⌣ — ⌣ ⌣ ⌣ — ⌣ (in this case one is justified in speaking of a resolution) has +678, being one of the two highest quotients of all.

^{page 237 note 2} Cf. Heymans, Zcitschr. f. angew. Psych. L. 323–4 and 366–81; William Stern, Die Differentielle Psychologie in ihren methodischen Griindlagen, 308–16; J. D. van der Plaats, Pharmaceutisch Wtekblad, August, 1906.

^{page 238 note 1} To this may be added the question of the resolutions and bye-forms.

^{page 238 note 2} Interesting results might be obtained in this way.

^{page 238 note 3} I hope shortly to discuss the investigations of Thumb (Fortschritte der Psychologie I), whose method seems to me to be capable of improvement, and whose results are not always to be depended upon.