^{page 364 note 1} I was going to add ‘nurses,’ as one would in speaking of any modern tongue; but then the nurses of Thucydides and Xenophôn would very likely be barbarians, and in any case no Athenian citizens. Indeed I ha forgotten the chance that Thucydide may have had a Thracian mother.

^{page 368 note 1} Besides the famous specimen of ‘Romana lingua’ in the ninth century in Nithard, iii. 5, I have collected several instances of the use of ‘lingua Romana’ or ‘Gallica’ (Norman Conquest, i. 618), in the ninth and tenth centuries. It had become so distinct from Latin that Latin was interpreted into it. In the eleventh century French literature begins. But in the twelfth Frederick Barbarossa speaks Latin as well as German. The ‘lingua Romana’ of Italy—where I think the phrase is not used—was still not recognized as a separate tongue, any more than any Romance tongue was in Charles the Great's day.

^{page 373 note 1} At a later time Greek had to strive with the Romance languages also; but the time with which we are now concerned is that of the earlier Eastern Empire, before the Romance languages had shown themselves.

^{page 376 note 1} Look for instance at the Novels vii. 1, (We hear in this novel of ἡ πρεσβῦτις Ῥώμη). Or again, xiii. 1, ἡ μὲν γὰρ πάτριος φωνὴ prœfectos vigihtm . xiii. 2, He talks of οἱ πάλαι Ῥωμαῖοι (just like Prokopios), ἡπρεσβυτέραῬώμη. Some of these are referred to by C. F. Weber, Dissertatio de Latine Scriptis quæ Græci veteres in Linguam suam transtulerunt, Cassel, 1852, ii. 35. Weber's subject is different from mine; but he has collected a great deal that incidentally bears on mine.

Later legislators seem to have been less scrupulous; that is, as Latin went out of use, such Latin words as could not be avoided became part of the Greek language.

^{page 379 note 1} There are few things more curious than the directions given by Constantine in his other work De Cœremoniis Aulœ Byzantinœ, (i. 74–77), for the cries, Greek and Latin, with which the Emperor is to be greeted by such and such persons at such and such times, as, for instance, One place is most curious, as showing a distinct misunderstanding of the Latin.

We have here got among those with whom

Non aliud est vivere quam bibere.

The beginning of the error is when we see Bixsit on tombs for Vixit.

Some of the ecclesiastical ejaculations are yet more amazing.

It is an odd fate for a Greek word like βαπτίζειν to become Latin, and then to come back again with its Latin inflexions in Greek letters. The light that all this throws on pronunciation is valuable. Mark how the Latin accent is kept in words like ὀμνήποτενς and δόμηνονμ.

^{page 382 note 1} I refer to the volume and page of the Bonn edition. References to the chapters in Prokopios often cause needless trouble, on account of the great length of the chapters.

^{page 382 note 2} This spelling is very important. Prokopios can hardly have lighted on such a Latin spelling as rœferendarius. It is plain that ε and αι had already the same sound, as we so often see in epitaphs Kite for κεῖται.

^{page 383 note 1} The use of bandus is illustrated by a passage of Stephanos of Byzantium, which is a beautiful case of heaping blunder on blunder. He is explaining the name Alabanda. (Such Romans, one would think, must, like some people in Herodotus, have sacrificed to a Karian Jupiter). But after all Stephanos is not so bad as the modern philologer who found Teutonic lady and Romance dame in some scrap of Lykian, Karian, or Alarodian.

^{page 384 note 1} Whence the h? The real Latin is triumpus, as Cicero witnesses (Orator, 48). He used to spell it triumpus, but changed his spelling because he heard everybody say triumphus. Why did they take the trouble? for I suppose he means phus and not fus, quite another sound. Then what is triumphus? Is it, as some think, only θρίαμβος after all? How then did a Greek word get into Latin so early? But I am concerned only with the fact that Prokopios so strongly identified θρίαμβος and triumphus that he could put the Greek word into Old-Roman mouths.

^{page 385 note 1} It might be worth inquiring whether Καρχηδών, as the name of the Roman colony, was ever anything more than a piece of fine writing, like calling Dyrrhachion Ἐπίδαμνος. I have to thank Mr. Bywater for sending me to a passage of Eusebios (Hist. Eccl. x. 6) where a letter of Constantine the Great begins, Long afterwards Nikephoros of Constantinople (84) has in a manner to define —would his readers have better understood him if he had said Καρταγένη?

^{page 387 note 1} Audite, omnes fines terræ, orrore cum tristitia,

Quale scelus fuit factum Benevento civitas

Hludowicum comprenderunt, sancto, pio, Augusto.

^{page 387 note 2} My Servian and other Slavonic friends must be greatly displeased at this bit of derivation. Some of them wish me to change my spelling of the name of their race, Slave, as I have spelled it from my youth up, and would have me take to the ugly fashion of the newspapers, Slav. Yet I cleave to Slave, if only to show, not that Σκλάβοι or Σέρβλοι were necessarily servi or δοῦλοι, but that servi or δοῦλοι came to be called slaves in Latin, Greek, Teutonic, and Arabic, because they were largely Σκλάβοι.

Ducange throws no further light on the question of shoes. The change from σ to τζ is strange anyhow.

^{page 388 note 1} The λογγῖνοι are said to be the same as λογγάρχαι. The derivation seems to fluctuate between λόχος and λόγγος. This last, which is explained as equivalent to κλεισοῖρα, seems to be the same word which we meet in Μεσολόγγι. My modern Greek dictionary does not acknowledge it.

^{page 389 note 1} I suppose it is a true etymology which connects νερό with Νηρεύς. Anyhow νερό has quite routed ἵδωρ.