intricacies of this subject. It seems to me that the liturgy referred to, differed from those of Basil and Chrysostom exactly in that part which is so carefully regulated by the council of Laodicea. First, we have no account of any prayers made after the sermon for catechumens and penitents, during the earliest ages of the Gallican and Spanish churches, who used the same liturgy. They are not mentioned by Gregory of Tours, by Isidore of Seville, nor by any of the authors referred to by Mabillon, in his treatise on the Gallican liturgy. It is true that Martene understands the author of the tract which he has published, and which is referred to Germanus bishop of Paris, to speak of prayers for catechumens. But I confess, that in perusing the passage to which he alludes, I am unable to see that the author of the tract does more than refer to prayers for catechumens, as an ancient custom of the church, which he seems to describe as a thing not then practised, and which he does not affirm to have been used in the Gallican church. Supposing, however, that these prayers for catechumens were used in Gaul, of which we have no sort of authentic evidence, they might very probably have been imported from the east about the same time as the Trisagios, which is prescribed in that tract, and which was first introduced into the eastern liturgy in the fifth century'. Secondly, we have no account of any prayer or prayers "of the faithful," in

See Martene, Thesaurus Anecdotorum, tom. v. p. 94.

f See Martene ut supra, p. 91, and Goar, Rituale Græcum, p. 126, not. 80, in Liturg. Chrys. The hymn Trisagios

mentioned here, must not be confounded with the seraphic hymn, or Tersanctus, which occurs in the course of the solemn thanksgiving, before the consecration.

the Gallican liturgy. But instead of this, the elements were placed on the altar, and the diptychs, containing the names of the living and dead, recited; after which the priest made a prayer. The kiss of peace and oblation succeeded, as in the liturgies of Basil and Chrysostom.

It seems, therefore, that the Gallican liturgy differed from those just mentioned, exactly in that part which is regulated by the canon of Laodicea. And since we should be able to account for that canon, by supposing that a liturgy like the ancient Gallican, prevailed in Asia and Phrygia before the council of Laodicea; and further, since we have independent reasons for thinking that the Gallican liturgy was derived in the second century from that of Asia and Phrygia; it seems to me highly probable that a liturgy which resembled the ancient Gallican prevailed in the exarchate of Ephesus until the fourth century, when it was altered by the council of Laodicea, in order to make it conformable to the great oriental rite, which has been used there ever since.

SECTION VI.

LITURGY OF THE PATRIARCHATE OF ROME.

IT has been much debated among learned men, whether the Roman liturgy can justly claim any considerable antiquity. Some have referred its composition to Gregory the First, commonly called "the Great," patriarch of Rome, A.D. 590. Others think it impossible at this day to ascertain the text, even as it stood in the time of that prelate 2. This subject has been confused, by mistaking for each other two very different things, the missal and the liturgy. The Roman missal (formerly called Sacramentary, or book of sacraments) was a large volume containing a number of missæ or offices for particular days, which were to be added, in the proper place, to the canon in which the more solemn prayers and the consecration were contained. By the Roman liturgy I understand the canon which did not vary, and the number and order of the prayers which were to be added from the missal.

a

Dupin, Hist. Eccl. cent. vi. tom. v. p. 102. Brett's Collection of Liturgies, p. 333, &c.

The various readings of manuscript sacramentaries are supposed to render it impossible to determine the text as it stood in the time of Gregory; but on examining these difficulties, it will be found that they do not prevent us from ascertaining the liturgy for all the variations, interpolations, and uncertainties of these MSS. relate to the individual missæ. In these I readily admit that a great variation, both of words and sentiments, may be found; and it is therefore a matter of some difficulty to decide which of the missæ are as old as the time of Gregory. Such doubts and difficulties, however, do not extend to the number and order of the prayers in each missa, nor to the canon. On the contrary, we find in all the same number of prayers, arranged in the same order, and designated by the same titles. The canon, or invariable part, preserves the same text in all MSS. The only difference that occurs is the introduction of some short prayer, or of the name of some person to be commemorated: but such interpolations are very rare, and when found are easily detected; and in no case is the canon itself either mutilated or altered. We can therefore ascertain both the invariable and the variable parts of the Roman liturgy. This agreement of MSS. in one common order and text derives strength from a consideration of the different ages and countries in which they were written. Manuscripts of Italy, of England, Germany, and Gaul, whether written at the same period or not, all furnish the same order of prayers and canon. To this evidence we may add the writings of various liturgical commentators in the eighth, ninth, and tenth centuries; which, though composed in different countries,

all concur in establishing the same facts as the manuscripts.

It appears, then, that there is no difficulty in ascertaining what the Roman liturgy was in the time of Gregory the Great. It may however be inquired, whether Gregory is to be considered the author of that liturgy. To answer this question, we must have recourse to ancient history. We are there informed with minuteness of the amount of Gregory's alterations and improvements. He collected, arranged, improved, abbreviated the collects of the individual missæ ". He inserted a short passage (which is known) into the canon. And he joined the Lord's Prayer to the canon, from which it had previously been separated by the breaking of the bread. All this amounts to positive proof that Gregory was the reviser and improver, not the author of the Roman liturgy.

An attempt has been made to prove that the Roman liturgy was composed between the time of Vigilius and Gregory. The former, who lived fifty years before the latter, speaking of the canon, said,