calling forth some of the most pathetic and touching epitaphs which can be found anywhere, that the death of those who had reached old age was regarded, not as caused by the cruelty of fate, but as brought about by natural law, or by the kind hand of Providence.

In C. I. L. VI, 4, Fasc. I, we meet with the following oftquoted epitaph: —

D.MM. VLPIVS CERDO | TITVLVM. POSVIT | CLAVDIAE · TYCHENI CONIVGI KARISSIM | CVM QVA VIX ANNIS | ... II. MENS VI. DIEB | III. HOR. X. IN DIE | MORTIS. GRATIAS | MAXIMAS. EGI | APVT DEOS · ET | APVT. HOMINES

Orelli includes this in his collection of Latin inscriptions (II, 4636). His only comment on it is the exclamation "mirum dicterium," but this clearly shows the meaning which he attaches to the words. Prof. A. Zimmermann, in an article entitled "Der kulturgeschichtliche Werth der roemischen Inschriften," after speaking of the genuine grief which monuments raised to husbands and wives so generally display, adds (page 9): "Nur eine Inschrift unter so vielen ist in einem unpassenden Tone abgefasst, es ist eine der Stadt Rom. Hier sagt der ueberlebende Mann: 'Am Tage ihres Todes habe ich meinen tiefsten Dank ausgesprochen vor den Goettern und den Menschen,' nachdem er sie kurz vorher seine theuerste Gattin genannt."

Friedlaender, who has devoted more attention to inscriptions than has any other writer on Roman life, in referring to women of the lower classes in Rome says ("Sittengeschichte," I, 516, ed. 6): "Nur Grabsteine von Frauen dieser Staende sind erhalten, auf denen ihre hinterbliebenen Gatten ihre Tugenden ruehmen; einmal freilich gesteht auch ein Witwer mit naiver Aufrichtigkeit in der Grabschrift seiner Frau: An dem Tage ihres Todes habe ich bei den Goettern und den Menschen meinen Dank bezeugt.'"

This epitaph, though regarded as an unparalleled exception, may be compared with V, 1, 3122, which might as readily suggest a tone unfriendly to the dead. This monument bears the names of two wives. To the first the husband applies the expression "uxori sanctissimae post obitum" in contrast to the second, who is addressed as "coniugi carissimae." A similar interpretation applies to both epitaphs, and though they are somewhat awkwardly worded by those who raised these humble monuments, still they contain no element of satire. In fact, though the Roman thought the tomb not an improper place to record plain truths about the dead, even though these truths were not always complimentary, and even to indulge in puns and jest, still he never displays the spirit of satire or ridicule. There may be words of indignation 1 occasioned by the ingratitude of the one to whom the monument was raised, or pity 2 for the weaknesses of the departed, but epitaphs, unlike all other departments of Roman literature, bear no trace, it seems to me, of the element of satire.

In the case of the epitaph under consideration it would indeed be strange, and with Orelli we should exclaim, "mirum dicterium," if the husband, after inscribing the epitaph to his "coniugi carissimae,” and after mentioning the length of their married life even to the hour, that which is extremely rare and which almost in itself implies that every hour of their married life had been dear to him, had in the next sentence thanked heaven that she was dead. Of course no one would deny that we must interpret the language of inscriptions by the usage of inscriptions, that we cannot expect in these epitaphs composed by the illiterate the logical clearness which characterizes Roman literature as a whole. We here meet with a boldness and license and a lack of propriety in the use of language which we do not elsewhere find. Expressions which, judging from the form, would seem directly opposed in thought are used to convey a similar meaning. For example, in VI, 2, 10703, we meet with the

1 C. I. L. VI, 3, 20905.

2 C. I. L. XIV, 636.

words "filiae pater non merenti feci"; though the literal meaning of these words is the opposite of 10696 of the same volume, "coniugi bene merenti fecit," it is intended to convey a similar idea. Again, in VI, 2, 6686, we find "fecit libes animo," and in VI, 4, Fasc. I, "titulum tibi feci libenter," but the meaning is not the opposite of the "dolens posuit" of XI, 557.

Though the writer of the epitaph under consideration has perhaps expressed himself awkwardly and with too great brevity, yet he used language whose meaning could hardly be misunderstood by his fellow-countrymen who were familiar with the usual brevity employed in epitaphs. The meaning appears perfectly clear to us when we compare the language here used with similar expressions which set forth the same thought more fully. C. I. L. VIII, Supp. I, 13134, is an epitaph inscribed by the wife in honor of her husband. The last sentence is as follows: "Sed ago superis gratias, quod, dum e[g]o viver[em], nil voluptatibus meis negavit, quia et ipsa meruera (m)." In C. I. L. VI, 4, Fasc. I, 29186, the husband returns thanks to his wife, "cui semper gratias." Here the verb is omitted. In VI, 2, 14537, the form of expression used is, "ago memoriae vestrae gratias." X, 1, 3162, is a fragmentary inscription which according to the restoration of Mommsen, and there can be little doubt with regard to the general correctness of his interpretation, reads as follows : cui ma[ritus] in die [funeris pia men]te grat[ias di]cit."

The epitaph we are considering differs from these last quoted simply in omitting some such expression as "memoriae tuae" or "pia menta," or some clause stating the reason for rendering thanks which would remove all obscurity even in the outward form. Other epitaphs of similar import might be added, but those already quoted are doubtless sufficient in number to remove any doubt with regard to the meaning which the writer intended to convey. We must accordingly acknowledge that this epitaph was intended as a genuine tribute of love and that its tone is far more pleasing than that of IX, 5813, "quod fas non fuit monimentum feci: quod inprecabo (or increpabo) superos et iferos." These two

epitaphs illustrate the two views presented by Seneca (Ad Marciam de Consolatione XII, 1), and the one under discussion displays that attitude for which he expresses his preference: "Si confessa fueris percepisse magnas voluptates, oportet te non de eo quod detractum est queri, sed de eo gratias agere quod contigit."

III. On the Accent of Certain Enclitic Combinations in

Greek.

BY PROF. FRANCIS G. ALLINSON,

BROWN UNIVERSITY.

JACOB WACKERNAGEL, who has done so much for our knowledge of Greek accent, contributed in a Baseler Programm for 1893 certain "Beiträge zur Lehre vom Griechischen Akzent." While Hirt in his recent (1895) Handbuch des Indogermanischen Akzents accepts some of Wackernagel's contentions, he rejects his ingenious explanation of the retraction of the accent in eywye, etc., and attempts to include this also under a new and, as it will perhaps seem to many, artificial formula which he uses to explain the shift of accent like that in μýτηp and μnτpós. Hirt's formula (p. 32) is: "Ruht der Ton auf einer langen Ultima, so wird der Akzent zurückgezogen," e.g. vos as against Skr. vasnás. He therefore assumes that an original *eyo and *po were thus preserved in the case of the nom. and dat. of eywye. Wackernagel's contention (Beitr. p. 20) was that the retracted accent of the stem euo- is older than the oxytone éμos; that therefore eμoɩye is older than poí; that eywye (although eyó corresponded to Skr. ahám, accent and all), developing on a combination later than eμolye, adapted its accent to the latter, while èμéye is explained away by assuming its development from *μeye and a prothetic e. As Hirt (p. 33) remarks, this is 'schön ausgedacht,' but is not convincing. But neither do I feel convinced of the value of Hirt's general formula, and consequently do not accept his incidental settlement of this point. The explanation previously (1891) suggested by me in a foot-note to p. 50, A. J. P. Vol. XII., is, I still think, as plausible as any. In arguing there against Professor B. I. Wheeler's brilliant