some change. Sound elements became more compactly united within syllables. The early diphthongs, e, ov, o, for example, though still written the same, had become monophthongs in the Hymn to Apollo (third century B.C.), as found inscribed on stone with musical notation at Delphi. And, again, certain groups of consonants finally came to be sounded, not partly with a preceding and partly with a following vowel, but all with the latter. This increased the number of open syllables, which meant increasing the ease and rapidity with which the language could be spoken. See, for example, the mute and liquid usage; such a combination in Homer generally causes length by position, but in the Attic poets generally not. In other words, these sounds during the early period were generally divided, but in subsequent times they were generally both joined to the following vowel. The extreme and latest stage in such an evolution is to be seen in the French language, where open syllables are relatively very numerous. Baudry says, concerning this later stage in the evolution of languages, when more consonants than one tend to be united with a following vowel: La voix a, pour ainsi dire, fait son éducation, et l'articulation, devenue plus agile, n'éprouve plus aucune peine à prononcer d'un seul coup deux ou plusieurs consonnes (Grammaire Comparée, p. 13).

Usage in dividing an unfinished word at the end of a line of writing is looked upon by some as the cause and origin of (b). Others think (6) was derived from the division of words into syllables when accompanied by musical notes. This is the view held by Professor M. W. Humphreys, being partially presented in his article on the "Equivalence of Rhythmical Bars and Metrical Feet" (Amer. Phil. Assoc. 23, 157). It seems, however, possible that, besides this, another factor entered into the case. The idea underlying (a) is manifestly older than the idea underlying (6). The former sprang from an early age of creative literary activity, being as old as quantitative poetry itself, while the latter dates from the subsequent age of criticism. And thus the language changes above described may not improbably have exerted some influence on those who formulated (b). Just how far (b) reflected actual usage of speech, it is now impossible to determine. However, even down to the end of the classical period the changes in pronunciation were not sufficient to induce either Greek or Roman poets to reform their theory of quantities in composing verses. Their poetry was therefore composed, and presumably always read, in accordance with the early method of word division. A slight disparity between the spoken language and the language as employed in poetry may have arisen in later times, but this is not to be wondered at. All peoples show greater conservatism in their poetry than in other modes of expression. We ourselves used to say, for example, wind, but gradually changed the word to wind, still retaining in our poetry, as the rhyme shows, the older sound wind.

We have seen that the rules are confused because they perhaps date from different periods. Another cause for their shortcomings is that, the science of phonetics being still undeveloped, they were based largely upon spelling rather than upon sounds. They often fail, therefore, to show how the sounds were produced or why the length resulted. Rule (a), for example, gives no idea whether one or both or neither of the "two consonants" was sounded with the preceding vowel. A third cause for complaint against the rules is that they often contain loose statements. See, for example, the rule for "common" syllables. It does

not make plain, first, that the necessary conditions for such a syllable are a succession of four elements, viz. short vowel, mute, liquid, and vowel (either long or short); secondly, that only a certain few combinations of mute and liquid by usage are here valid; thirdly, that the succession must be entirely within one word; fourthly, that if the word be compound, the succession is always entirely within one member of the word; fifthly, in case of a common syllable used as short, the division of sounds is v-mlv, but, in case of such a syllable used as long, the division is vm-lv.

The Chair having been informed of Professor Pease's inability to serve on the Committee on Nomination of Officers, Professor Murray was appointed to fill his place as chairman of that committee.

The Committee on Time and Place of Meeting being ready to report, that report was then called for. The committee recommended that the next meeting be held at San Francisco, on Friday and Saturday, December 28 and 29, 1900. The report was adopted, and the meeting adjourned at 12.45 P.M.

FOURTH SESSION.

The fourth session was called to order by the Chair at 2.15 P.M.

17. Faust-Interpretations, by Professor Henry Senger, of the University of California.

This paper is published in Mod. Lang. Notes, XV. 82 ff.

18. The Use of the Optative with ei in Protasis, by Dr. J. T. Allen, of the University of California.

No abstract of this paper is available.

19. Supposed Irregularities in the Versification of Robert Greene, by Professor C. M. Gayley, of the University of California.

The author contended that Greene's dramas were written not to be read, but spoken, and that a proper observance of various rhetorical pauses would show that most of the apparent irregularities in the verse were intentional and highly artistic.

The paper will appear in Vol. I. of the author's Representative English Comedies, Macmillan, N. Y., now in press.

20. The Potential Subjunctive in Latin, by Professor E. M. Pease, of the Leland Stanford Jr. University.

After considering the meaning of the term potential and its usual application there followed a brief consideration of the views of Elmer and Bennett on the

subject. The former would eliminate the potential subjunctive entirely from our grammars, and the latter would recognize only a few stereotyped may potentials and one small class of can-could potentials. Exception was taken to Elmer's fundamental principle that "unless one instance of the subjunctive can be cited which cannot possibly be explained except by assigning to it the force of the potential, then there is not the slightest justification for claiming that the subjunctive has the power of expressing this idea."

A strenuous application of this method would eliminate many other important grammatical categories. Moreover, until it is proved that the potential idea was not expressed in the modal forms of the parent speech, the presumption is in favor of the potential subjunctive in Latin. The current view of the best authorities is for a modal potential in both ancient and modern languages. Therefore, to attack the potential in Latin successfully, one must at the same time eliminate it from all the related languages. For if we grant there is such a thing as a potential subjunctive in other languages, every unprejudiced observer must admit that many a subjunctive in Latin appears to yield its closest meaning only when translated by may, can, might, could, etc. This is true of many of the passages quoted by Elmer.

To approach the question from another point of view, there are in Latin and in other languages various synonymous constructions or interchangeable ways of expressing common ideas; e.g. commands may be expressed by the imperative, the subjunctive, and even the indicative. Various are the ways of expressing purpose; likewise cause, time, agency. The shifting of voice from active to passive, and vice versa, is but a matter of emphasis; necessity and obligation have different modes of expression, and so on with other categories. Is it then likely that the Latin language was so poor in expressions of power, ability, and possibility, that there was but one mode, the auxiliary verbs, posse, quire, etc.?

Elmer asks why we do not find many instances of verum sit, pluat, etc., in Latin if the Romans made use of the subjunctive. The answer would be: for the same reason that verum esse potest, pluere potest, are rare; the ideas should not be expected frequently in the kind of literature that survives. The thought "it may rain" was doubtless frequently expressed, but after a careful search I have failed to find a single case of either pluere potest or pluat.

The Latin subjunctive is known to be a development from the indicative, — a sort of specialized or indistinct future. Where the indicative asks "what will you do?" the subjunctive inquires "what are you to do?"—with the time idea less definite and the contingent idea more distinct. This meaning of the subjunctive is fundamental, and found in nearly all cases of its independent use. In Cicero, Verr. 2, 16, “ Quid hoc homine facias?” the fundamental idea is “what are you to do with this man?" This permits of more definite translation if a protasis is expressed. Thus if the context emphasizes the idea of contingency with a conditional protasis, then we translate “what would you do, etc.?" If the idea of power, ability, is present, then we translate “what can you do, etc.?” A temporal protasis often shows the independent clause to be a can or could subjunctive; an adverb often indicates a may or might subjunctive. Sometimes these subjunctives are purposely indistinct and colorless. At any rate the essential qualities of the mood do not change with the protasis, whether expressed, implied, or purely elliptical. A careful search will discover a goodly number of

potential subjunctives in all periods similar to the following: Juv. 3, 112, Despicias tu forsitan inbellis Rhodios unctamque Corinthon, despicias merito, you may perchance despise. and justly too; Cic. Pro. Rosc. Amer. 89, Ego forsitan in grege adnumerar, as for me I might perhaps be counted in the common herd (Lane 1556); Cic. Pro. Planc. 64, Vere, mehercule, hoc dicam, surely this I can indeed say.

Sometimes the can-could subjunctive is used in close promixity to posse, licet, or an adverb suggesting ability: Verg. Ec. 1, 40, Quid facerem? neque servitio me exire licebat, what could I do? It was not in my power, etc. Liv. 21, 4, 3, Itaque haud facile discerneres, utrum imperatori an exercitu carior esset, so you could not easily discover whether he were more beloved by the commander or the army. The potential is common also in relative and other subordinate clauses. Plaut. Ps. 294, Nullus est tibi, quem roges mutuom argentum? have you no one you can borrow of? Liv. 21, 36, 4, Haud dubia res visa, quin per invia circa nec trita antea quamvis longo ambitu circumduceret agmen, there seemed to be no doubt the army must be led around, etc.

The report of the Committee on Nomination of Officers for 1899-1900 was then called for. Professor Murray reported as fol

lows:

President, Benjamin Ide Wheeler, University of California.
Vice-Presidents, Ewald Flügel, Leland Stanford Jr. University.

E. B. Clapp, University of California.

Secretary and Treasurer, John E. Matzke, Leland Stanford Jr. University.
Executive Committee, The above-named officers and

E. M. Pease, Leland Stanford Jr. University.

W. A. Merrill, University of California.

Julius Goebel, Leland Stanford Jr. University.
C. M. Gayley, University of California.

It was voted that the Temporary Secretary be instructed to cast the ballot of the Association for the persons named in the report, whereupon they were declared duly elected.

Upon motion of Professor Goebel it was then

Voted, That the Association tender a vote of thanks to the authorities of the Mark Hopkins Institute of Art for the use of the room in which the sessions of the Association were held.

There being no further business, the Chair then declared the meeting adjourned.

INDEX.

Arabic numerals indicate the pages of the Transactions; Roman numerals indicate the pages
of the Proceedings.

Amalgamation, illustrated in Roman re- capere for accipere, 204.

Catullus viii, interpretation of, xxxix;
lxxiii, Attis in, 46, 55, 56.

Antigonus of Carystus, used by Pliny, cedere for excedere, 206, 211.

Aristophanes, early literary history of, Chorus in New Comedy, its relation to

Athenian democracy in the light of Cicero's use of imperfect and pluperfect

lxvi