1920 REVIEWS OF BOOKS 587 the change, especially in Scoto-British circles, whereby the benefit of absolu- tion became almost secondary to the subjective value of self-examination and confession. The author is over-sceptical of the use of penitentials in papal Italy from 650 to 950, and he rather explains away than explains the references to a Poenitentiale Romanum ; but there must have been some treatment of sins that did not demand the greater excommunica- tion. May not the zeal for uniformity which led Charles the Great to prescribe the Gregorian Sacramentary as the official mass-book have also, through Alcuin and other English scholars at his court, been the cause of the adoption of the British use of penitentials ? It seems needless to transcribe the reference to Hernias in the Muratorian Fragment in the form nuperrim et tem'poribus nostris, when the diorthota has corrected it to nuperrime; on the other hand fine is printed where the stroke above the line shows that fi>nem was meant. Surely Hermas's message to the apostate Maximus is ironical : ' if it seem good to thee, deny Him again : is it not written, in your precious Eldxid and Modat, " The Lord is nigh to them that turn to Him" ? ' It is incredible that Hermas, who never cites a scriptural book, should attach authority to this dubious work. On p. 26 the same extract is printed twice under different headings ; and on p. 636 ' laici cum coniugibus et familiis suis ' is oddly paraphrased
- married persons ', as though bachelors were excused from penance.
The corrected proof appears to have been entrusted to a compositor who was unacquainted with Greek, for so only can we account for the astounding misplacements of accents and breathings, especially over diphthongs, the confusion of enclitics, and similar errors which disfigure every page of Greek. Other languages fare better, though c and e are confused in ' planetu ' (p. 93), ' ci ' (p. 94), ' mercretur ' (p. 98), ' die ' for ' die ' (p. 372), ' luetu ' (p. 377) ; we also find ' indique ' for ' indigne ' (p. 145), ' guadere ' (pp. 372, 376), ' veretur ' (p. 376), ' horam ' for ' norani ' (p. 404), and ' ham ' for ' nam ' (p. 524), ' epsicopis ' (p. 574), ' then ' for ' than ' (p. 74), ' heigen ' for ' heiligen ' (p. 231) — a list which is by no means ■ exhaustive. War reminiscences are presumably responsible for ' a cantena MS.' (p. 318). The careful table of contents makes an index almost unnecessary, but it must be observed that the index supplied is wholly inadequate. A. Bevil Browne. Les Origines de VAncienne France. X^ et XL Siecles. Tome iv, Les Nationalites Regionales. Leurs Rapports avec la Couronne de France. Par Jacques Flach. (Paris : Recueil Sirey, 1917.) In reviewing M. Flach's third volume, published in 1904, a statement was made of the general position taken up by him which need not be repeated here.^ Some doubt was then suggested as to whether, despite all M, Flach's learning and ingenuity, he had demonstrated his paradox that early France was no ' feudal ' state, and that the tie which bound the ' princes ' of Gaul to the early Capetian monarchy was not ' feudal ' vassalage but a vague ' supremacy ' of the over-lord, such as, one imagines, we find in the Anglo-Saxon system from a very early stage onwards. The full ^ Ante, XX. 141-3.
