remoter period represented the second and first grammatical persons reflexively. At the time written language brings se to notice it is in a worn-out condition. It retains no mark of gender or of number. Even the genitive case has vanished. In actual use the same inflexions of SE are, according to the context, considered either as singular or plural. It is highly probable that in its earlier condition the reflexive se was fully inflected for all relationships customarily represented by inflexion. Logically there is necessity for reflexives in all three persons. And if there were some primary sound sa there can be no a priori reason why number, gender, case, and person should not have been represented by developments of this radical. At any rate it is now admitted as a working hypothesis that verto-r is the same as verto-se, and on this hypothesis, as verto se means “I turn myself,” se here represents the first person. On the general question that the reflexive verb precedes the passive, that in the growth of verbal forms the middle or reflexive verb is historically antecedent to the passive forms, there has been since Bopp's time substantial agreement. Bopp (ii. 648) enunciates his view briefly thus: Ma-mi, sa-si, ta-ti, are suffixes naturally formed by reduplication. If, then, ma signifies me, ma-mi signifies myself. By parity of formation, sa-si, ta-ti, mean thyself, himself. Hence arise the suffixes of the present indicative reflexive of the Greek verb. Ma-mi falls away into-μαι, sa-si into -σαι, ta-ti into -ται. Bopp points out that in Old Slavonic the Accusative of the reflexive pronoun is added to the transitive verb to give it a Reflexive or passive significance. He illustrates from Lithuanian, which attaches the consonant -s without vowel mediation to the active voice to form the reflexive verb; under certain conditions also it prefixes the reflexive pronoun with the same result. In 1846 Key advanced this view of the Latin middle voice, as being not an application of the passive voice, but as being the actual forerunner of the passive voice. Key said (Lat. Gr., 2nd edit., p. 59),— “In Latin a reflective suffix is added to a transitive verb, so as to give it the reflective sense. “A reflective verb thus denotes an action upon oneself, and in Latin is conjugated in the imperfect tenses with a suffix -s or -v. An intransitive verb is generally in meaning reflective: as cur—i.e., put oneself in a certain rapid motion; ambŭla—i.e., put oneself in a certain moderate motion; but, as the object in these cases cannot easily be mistaken, no reflective pronoun or suffix is added. “When the source of an action (i.e., the nominative) is not known, or it is thought not desirable to mention it, it is common