¹ For a brief statement of some of these see Körting, Lat. Rom. Wörterbuch, 3d ed. (1907). For a more complete presentation see Georg Stucke, Franz. Aller und seine Verwandten, Heidelburg diss. (Darmstadt, 1902).

² Except P. Marchot (Studj di Fil. Rom., viii (1901), 391), who presents it as if it were brand new.

³ See Stucke, op. cit., s.v. ∗allare.

⁴ But less completely than is often supposed. Offrir is a derivative of obferre. In the twelfth century St. Bernard uses profiere for modern préfère (cited by Littré s.v. préférer), and its diphthong shows it to be a popular word. The stem lat probably survives in blé (oblatu), and in some common Church words, learned of course, e.g., prélat, oblation, etc.; while in Spanish we have the popular form trasladarse, which is very close to our meaning of se adferre, as Baur brings out. Other more or less popular survivals of this supposedly dead verb might easily be discovered, aside from such words as préférer, transférer, déférer proférer and others of the type.

⁵ See Stucke, op. cit., s.v. ∗ambitare; also L. Wiener, Byzantinisches in Zs. für Rom. Phil., xxxvi (1912), 385 ff., esp. p. 407; ibid., xxxvii (1913), 569 ff.

⁶ It thus appears earlier than any Romance andare. Wiener (Zs., xxxvii, 569 ff.) says that Latin andare is common in S. Italy in the early ninth century, but his earliest specific citation is dated 956.

⁷ Ambulare in the basic meaning “go” may likewise be called extinct, for in Old French it means only “amble,” and in Rumanian it means “travel,” “wander,” “circulate.”

⁸ Romania, ix, 624. His own choice, addere giving ∗addare, shows the same prefix.

⁹ Se auferre, “leave,” “get out” is both pre-classic and classic: “auferte domum, abscede hinc,” Plautus, A sin. 2, 4, 63; “te obsecro, hercle, aufer te modo,” Plautus, Rud. 4, 3, 93; “seque ex oculis avertit et aufert,” Virgil, A en. iv, 389.

¹⁰ In his Lat. Rom. Wörterbuch, s.v. ambulare.

¹¹ For the relation between active, passive, reflexive, deponent, and middle forms of verbs see Kühner, Ausführ. Lat. Gram., i, 649–653.

¹² I use the term deponent as applying to a verb which has developed a passive form, originally in normal relation with the active, but later separated from it. It comes to express an active idea in line with its former passive one (e.g. from a form ∗morere, “kill,” mori, “be killed,” “die”), and may exist separately but simultaneously with the active form, or may outlive the latter.

¹³ It also survives in the transitive sense “take out,” whereas in France partir, in the simplex, lives only in the deponent sense.

¹⁴ Förster and Koschwitz, Altfr. Übungsbuch, 5th ed. (Leipzig, 1915). See the Nachträge for bibliography.

¹⁵ As cited above. He considers aller a development of andare, first found in a document of 956. He reasons that much time must have been required for this development, hence, inter alia, he derives grounds for his late dating. He does not mention the fact that aler exists already in ca. 900.

¹⁶ This fact is recognized by Kurt Hetzer, “Die Reichenauer Glossen,” in Zs. für R.P. Beiheft vii (1906); he also recognizes the other two similar pluperfects. He attributes alaret to ambularat.

¹⁷ No. 275, sugg(ess)erat: dixerat, ortaret; and 886, diferbuerat: exbuliret.

¹⁸ In six cases the preterit is expressed unambiguously by this pluperfect: auret (2), pouret, furet, voldret, roueret. Three forms are ambiguous (perf. or pluperf.): voldrent (2), getterent. The perfect is used in fut (2), but contredist (?), volt and volat are ambiguous (present or perfect). The historical present is also used: eskoltet (?), ortet, ruouet. See Schwan-Behrens, Gram., 337, 2a.

¹⁹ Ed. Gaston Paris, 6th edition; see vocabulary.

²⁰ It may have had several meanings at the same time, just as in modern Spanish it is imperfect subjunctive, preterit, and pluperfect indicative, and also conditional.

²¹ In view of the ultimate e we should expect free accented a to appear also as e if that change had occurred when the word was recorded, since the ultima indicates northern Gaul (certainly not the south) as the source of the document—provided it has a single source. Gloss 275, sugg(ess)erat: dixerat, ortaret, with its classical dixerat and Romance ortaret, seems to show that the entries were not all made by the same person.

²² I now venture to omit the traditional asterisk, for alaret has every indication of complete popularity, and the verb may be considered as attested.

²³ During the transition period, when deponent verbs were becoming active in form, it probably had two perfect forms, aliavi and allatus sum, identical in meaning. There must have been such doublets and triplets for the same tense as mortus sum-mori, ausus sumausavi-habeo ausatus, usus sum-usavi-habeo usatus.

²⁴ As defined in Note 12.

²⁵ Old French forms in -g- (alge, auge, etc.) are analogical to the same obscure model which gives similar forms in -lg-, -rg- and -ng- (donge, moerge), found in early literature of the western dialects. That this -g- is phonetic, and not part of a palatal graph, is shown in the rhyme of “N'i remaint dame qui n'i vienge / Las! ja n'en tornerunt mais, ce crien-ge,” from Benoit, Chron. des Ducs de Normandie, ed. F. Michel (Paris, 1836), i, 1681–1682.