type appears. Rude as it is, the hermit's chapel or oratory 1 on Inchcolm possesses features in the radiating vault of its roof, its grouted and squared-stone covering, the arching of its doorway, its position, and even in the approximately quadrangular form of its ground-plan, which will all take rank as features of advancement after I have shown (as I hope to do in my next lecture) what were the characteristics of the earliest forms of structure consecrated to the service of religion when the church was first permanently planted in

1 In dealing with these structures by archæological methods, it is not necessary to observe ecclesiological distinctions that may exist regarding the precise application of such terms as "oratory," "chapel," or "church.' Wherever they are used throughout these lectures, they are used without reference to ecclesiological limitations, and the word "church" is employed as a generic term, embracing all the varieties of the type of structure designed for Christian rites. Dr. Petrie, speaking of the early stone-roofed structure on Cruach MacDara, which measures 15 feet by 11 feet internally, calls it sometimes the church, and sometimes the oratory of St. MacDara. Teampull Cennanach, measuring 164 feet by 124 feet internally, he also terms indifferently an oratory and a church. In fact there is no possibility of applying the ecclesiological distinctions between an oratory and a church to the actual remains, because it is a distinction founded on the ancient use of the edifice of which there usually exists no record. It might be otherwise if there could be a distinction drawn from existing characteristics, such as from dimensions or architectural features; but while it may be possible to say that a singlechambered structure of the Christian type, which exceeds 30 feet in length, is not likely to have been an oratory (in the ecclesiological sense), it is not possible to say that a very much smaller one may not have been a church. On the one hand we read in the Irish Annals of an oratory in which 260 persons were burnt; and on the other we know that the small cell called Teampull Ronan, on North Rona (described in the next lecture), was the church of the islanders as long as the isle was inhabited. It may have been an oratory in the ecclesiological sense, when the founder was the only Christian worshipper on the island, but in this view every primitive church erected by an individual founder in a Pagan district must have been originally in the same position. But even if it were possible to make the distinction, it is rendered unnecessary by the fact that such edifices as are known to have been "oratories" or "hermit's chapels" do not differ in their typical character from the smaller variety of single-chambered church.

Scotland. And to learn the special features of that earliest style of Christian construction we must look to Ireland, the ancient Scotia,1 where the genius of the people, their immemorial customs, their language and institutions, were so similar to those of our own country that when the new faith was finally established by the labours of her missionaries, the converts accepted with it the ecclesiastical customs, constitution, and usages already established there.

In this lecture I have traced the typical form of the twelfth century church back to the transition stage through which it passed out of the simpler form that preceded it. We have seen it associated with the Round Tower of the Irish type, and deduced from this and other indications that the Round Towers of Scotland, like the principal group in Ireland from which they are derived, are not the associates of the earliest types of the architecture of the Celtic church, but of the latest, that is of the type that passes directly into the decorated style and elegant construction of the Norman manner. We have traced the type of the chancelled church through various gradations back to a degree of rudeness and simplicity, which substitutes for the chancel-arch a flat-topped opening having its sides inclined towards each other instead of perpendicular, and differing in no respect of size or construction from the entrance doorway. We have seen the transition from the double- to the single-chambered form of structure by the addition of a chancel not bonded into the nave, and the primitive type has been finally reached in the small church of one chamber, one door, and one window. In

1 "The voice of all Antiquity pronounces Ireland to have been Scotia : To omit a host of authorities, Adamnan's Life of St. Columba and Bede's Ecclesiastical History ought to have been sufficient to prevent a question being raised on the subject." To this testimony of Dr. Reeves the reader may add that of W. F. Skene in the introduction to his Celtic Scotland.

the next lecture I shall trace this utterly simple form through farther gradations until it reaches the utterly simple character of construction that consists in the placing of stone upon stone without any binding material to keep them together, and becomes associated with fortified enclosures, and beehive-shaped cells, thus linking the Christian types of structure with other types which stretch back into purely Pagan times.

LECTURE III.

(21st OCTOBER 1879.)

STRUCTURAL REMAINS OF THE EARLY CELTIC CHURCH

Continued.

NEITHER the history nor the remains of the early Christian period in Scotland can be studied apart from those of Ireland. The ultimate establishment of the Christian Church in this country was the work of Irish ecclesiastics, and was therefore an extension into Scotland of the ecclesiastical system then prevailing in Ireland.

It follows from this that the study of the early Christian remains in Scotland is the study of a derived group, exhibiting local peculiarities, but possessing the general features and characteristics of the principal group of which it is an outlier or an offshoot. I might even go farther, and say that in Ireland itself the study of its early Christian remains is also (though not equally) the study of a derived group, inasmuch as Christianity did not originate there, and its adoption consequently implied the introduction of usages,—such as writing for instance; of styles of construction, such as building with lime; of typical forms of structure, such as churches and oratories; and applications of ornament, such as the carving of memorial crosses-which had no previous existence in the country. But it is sufficient in the meantime to indicate the principles on which the investigation must proceed. These are (1) That the typical characteristics of a group are most

readily obtained from the study and comparison of the greatest possible number of the most perfect specimens; (2) That this number is more likely to be met with in the principal group than in the derived group; and (3) That the characteristics thus obtained, as typical of the principal group, will also be present in the derived group in consequence of its subordinate character.

The earliest churches in Ireland were constructed within the fortified enclosures of the chiefs who embraced the faith, and took the founders under their protection. Thus we learn from the tripartite life of St. Patrick, that the church at Donaghpatrick, one of the earliest erected by him in Meath, was built where the house of Conall, the king's brother, was situated, which was given up to St. Patrick for the purpose. The church of Cill Benen was erected within the fortress of Dun Lughaidh, so called from a chief who with his father and four brothers was baptized, and gave up his Dun for the purpose.1 When Aodh Finn, the son of Feargna, was converted by St. Caillin he gave up to him his Cathair, or stone fortress, in order that he might erect his monastic buildings within it. The system which thus arose in the incipient stages of the church's growth, continued long after the circumstances which rendered it necessary had passed away. The association of the church with a fortified enclosure, which had been at first dictated by necessity, became established by long custom as the normal form of the ecclesiastical structure, and the rath or the cashel3 surround

2

1 Petrie, Ecclesiastical Architecture of Ireland, p. 444. In 1826, when Dr. Petrie visited Kilbannon, the remains of this great Rath, a portion of the circle, was still to be seen; in 1838, when Dr. O'Donovan visited it, all traces of the enclosure had been swept away.-Dunraven's Notes on Irish Architecture, p. 72.

2 The Book of Fenagh, quoted by Petrie, loc. cit.

3 The oldest forms of defensive structure mentioned in ancient Irish writings are the Caisel, the Rath, the Lis, the Cathair, and the Dun. The