he has quite correctly enlarged some of its indefinite entries.[783] His use of this martyrology is clear from his inserting Eusebius among the saints, in ignorance of his Arianism; he inserted the name in all simplicity, a mistake avoided by the Greek menologies, and the same may possibly have been the case with regard to Arius in the Weissenburg Codex.[784] This employment of earlier sources coupled with the numerous correct entries in the _Hieronymianum_ entirely excludes the view that the document is a fabrication. Indeed its composition can safely be said to have come about in the following manner:—in the two first centuries the persecutions were on the whole local, and the number of martyrs not very large, although the persecution at Lyons in 177 caused the death of more than forty martyrs, not counting confessors; a change took place in the third century, when persecutions were commanded by the emperors for the whole extent of the empire, and, under Diocletian, the martyrs were to be reckoned, if not by millions, at least by thousands. At the conclusion of the persecutions, it must have occurred to many to ask how many had lost their lives in this troublous period, and the idea of drawing up statistics of the martyrdoms must have sprung up. We have no evidence of anything of this sort actually having been done, but in the sixth century there was a widespread opinion that a work of this kind had been accomplished by Eusebius under Constantine, and the passages quoted from Cassiodorus and Gregory the Great show it was believed a list of the martyrs for the whole year existed in Italy or in Rome. If this work contained merely a list of names and dates, its interest must have been merely statistical and in no wise scientific, since little could be learnt from the names by themselves. In the seventh century, however, either in Auxerre or in North Italy, it seems to have been held in greater esteem; it found a transcriber and reviser, and finally was brought into connection with the liturgy owing to its being read at the choir-offices in monasteries and convents, and by this means, it won its way to a position of widespread importance, which, however, did not have an advantageous effect on the purity of the text. With regard to the printed editions, there are several published by various scholars who relied upon late MSS. of little value. We may mention the editions of Fiorentini, Lucca 1668 (incomplete), of D’Achery in the _Spicilegium_, ii. 1 (Migne, xxx.), and of Galesinius, Milan, 1577. It used to be disputed which of these editions gave the earliest text, but they are all quite superseded now. There are also many abstracts of the _Hieronymianum_ in mediæval handwriting, as, for example, the _Martyrol. Gallicanum_; Martène, _Ampliss. Coll. VI._, called after Chauvelin (Migne, _Patr. Lat._, xxx. 607); the _Fuldense_; _Anal. Boll. XIII._; the Reichenau martyrology, etc., but they are of small scientific importance.