ARCHITECTURAL REMAINS OF THE ROMAN
PROVINCES.

On the Architectural Remains of the Roman Provinces. By JAMES BELL, Esq.-(Paper read at the Opening Meeting of the Session 1850-51 of the Royal Institute of British Architects.)

ALTHOUGH the architectural remains of Rome and Italy in general have been thoroughly investigated, there is reason to reflect that other parts of the extensive empire of the Mistress of the World have not received the attention they deserve; and that while there is no doubt that a perfect and correct taste is only to be acquired by a study of the best works, it should not be forgotten that many remains of inferior artistic merit may yet present varieties of form and idea deserving of attentive consideration. The more complete our knowledge is of the whole scope of ancient art, the more likely are we to be correct in attempting to realise a just conception of the original nature of the best works, many of which are now in so hopeless a state of ruin. The student, therefore, might do well to add to his examination of the remains of the Eternal City itself, that of some of the Roman provincial towns, which the want perhaps of a correct appreciation of the character of that people has hitherto caused to be, in great part, neglected. Nothing marks more strongly this character in the later ages of the Republic and during the rise of the Empire, than the method pursued by them in dealing with surrounding nations. They had no desire for allies and tributaries; their treaties were hollow and treacherous, intended to be broken on the first opportunity, conquest being their only aim,-not with the view of obtaining subjects, but in order that each nation should in turn be incorporated with Rome, and form an integral part of the great empire. For this reason, submission was quickly rewarded with freedom and every Roman privilege, while resistance was punished to the last extremity. We should therefore naturally conclude, that evidence would be found in the monumental records of the provinces of Rome, of this identity with the mother country-and this we find to be the case. Germany, France, Spain, Africa, Egypt, and Palestine, all contain specimens of greater or less magnificence and grandeur, many in a high state of preservation, and some possessing peculiarities of form, construction, and arrangement, which render them highly interesting and valuable. I shall endeavour to call attention to some of the most important of these monuments, and to show that correctly measured representations of them deserve a place on our shelves, by the side of the illustrated works by Stuart and Revett, and by Taylor and Cresy.

Omitting all allusion at the present time to the underground remains which abound in this country and elsewhere, as well as to the roads and fortresses, which belong rather to engineering than architecture, we need not go far to find some monuments of considerable importance. The Porta Nigra at Treves is a stupendous work-two towers, 90 German feet in height, and more than 30 feet in diameter, and decorated with four orders of columns, are united by a curtain, 55 feet in extent, and three stories high, in which are the two gates; excepting the lower story, the whole is also arcaded. This design of a gateway is unique, and it exceeds in dimensions any similar building elsewhere. Treves possesses also the remains of an amphitheatre, in which some peculiarities of the substructure are apparent, also those of a basilica; but next to the gate, the monument of Igel is the most curious, and may be compared with one of a similar description at St. Remi in the south of France. There was also a monument of a similar description at Arlon, but the ornaments were transported in the 16th century by the Count de Mansfeldt, to form part of a collection of antiquities long since dispersed: that of Igel was happily preserved from the same fate. Had these monuments attracted the attention they deserve, some better idea could now be given than a mere verbal description. It was one inevitable consequence of the Roman policy before described, that all remains of a former civilisation were inevitably obliterated, but there are strong reasons for supposing that this neighbourhood was very far from being in a state of barbarism before the Roman conquest. We may believe this without going back to the early foundation of Treves claimed by some historians.

But it is in the south of France that we can more completely form a correct idea of a Roman province. This portion of France formed the province of Gallia Narbonensis, having been subdued from one to two centuries before the Christian era. Here the arches, gates, temples, amphitheatres, and aqueducts, rival those of Rome itself, which possesses no temple of the kind so perfect as the Maison Carrée, at Nismes. This edifice is a hexastyle pro

style temple of the Corinthian order, and originally stood in the centre of a forum, the extent of which has been traced by the bases of several of the columns found in situ. The Temple of Diana, as it is called, but which was in reality the Hall of the Baths, exhibits a beautiful arrangement of pilasters and niches in the interior, connected with a shrine in the centre of one end, of great elegance and originality. The amphitheatre, though smaller than the Coliseum, is in a far better state of preservation. This was built at the expense of Antoninus Pius, whose ancestors came originally from Nismes, and the Maison Carrée was dedicated to his adopted sons, Lucius and Marcus. It will be seen from the plan of the amphitheatre, that it is constructed somewhat differently from the Coliseum. We can here study the preparations for the velarium; the arrangements of the seats, galleries, and staircases. The podium round the interior is formed of single stones, 5 feet in height, to retain the water for the purposes of the naumachia.

The antiquities of Arles consist of an amphitheatre, a theatre with two columns of the proscenium still standing; and innumerable tombs and sarcophagi. Orange possesses a theatre of the most gigantic dimensions; the seats are cut out of the side of a hill, and the scene wall rises to a height of more than 100 feet by 300 feet in length. Though the marble decorations are, as in many other instances, almost entirely gone, it is still a most interesting relique. There are also several arches remaining of the hippodrome, and a beautiful triumphal arch in a very fair state of preservation. The ornaments have suffered from the singular purpose to which the building was appropriated by the Princes of Orange-when it was built into the Chateau, and the archway formed the principal salle de reception. The portion of the Roman aqueduct, now called the Pont du Gard, is too well known to require a detailed description.

The monopteral monument at St. Remi is of most beautiful design and proportions, and well deserves study as a model of this description of edifice. The gates of Nismes, Besançon, Sens, and Saintes, between La Rochelle and Bordeaux; the amphitheatre, aqueduct, and the Porte Dorée at Frejus, the birth-place of Agricola; the arches at St. Remi and Carpentras; the bridge and arches at St. Chamas, between Arles and Marseilles; and the innumerable fragments collected in the museums of Nismes, Arles, Avignon, Narbonne, and Toulouse, offer to the student who wishes to become acquainted with Roman art in the time of the Antonines, the strongest temptations to be found within the range of a summer excursion.

Spain offers an example no less striking of the peculiar character and vicissitudes of a Roman province. All vestiges of early civilisation previous to its subjugation are gone, and in its place we find most extensive remains of Roman enterprise and constructive skill, of which it is much to be desired that we possessed more detailed and illustrated descriptions than those which are at present within our reach. We learn, however, from the hand-book, that there is a Roman bridge at Merida, 2575 feet long, besides numerous antiquities, among which is a peripteral temple. At Alcantara is a bridge of Trajan, 600 feet long, and 245 feet above the usual level of the river. At La Barca five arches remain of a Roman bridge; the same at Capara; at Toledo there is a temple; near Tarragona a superb aqueduct, and a monument called the tomb of the Scipios, and at Segovia an aqueduct, 2500 feet long.

Passing now to Africa, we find that the same destruction of previous evidences of civilisation took place here as in other colonies. All that remains of Punic Carthage are a few inscriptions occasionally dug up: everything else is Rman. The best illustrations we have of these ruins are contained in two volumes of drawings by Bruce the African traveller, which are now in the royal collection at Windsor. Besides the usual amount of triumphal arches, some of which are of forms not elsewhere met with, there are other buildings of an unusual description. Of the first class is one large square inclosure at Suffetala, entered by a large triumphal arch, and containing three Corinthian tetrastyle temples connected together. At Lambesa a building, something in the form of a basilica, now roofless, entered on each of the four sides by a large centre arch, with two small ones flanking it; two orders of engaged columns, with broken entablatures, forming the exterior decoration. At Thisdrus, a very fine amphitheatre, approaching the Coliseum in size, and even surpassing it in state of preservation. Tripoli, an arch of the time of the Antonines. Considering the Cyrenaica as a Greek colony rather than a Roman province, we may omit a detailed description of the remains, which, we are informed by Captain Beechey, consist of sculpture of the best style,

At

with tombs, pavements, theatres, amphitheatres, and city walls, very perfect. Greece and her colonies, as well as Egypt, formed some exceptions to the usual routine of Roman conquest-Grecian civilisation acquired the respect even of the imperious Roman, and severity was only exercised towards the Grecian race, when provoked by imprudent resistance. Egypt also wisely submitted -Ptolemy bequeathed his kingdom as a legacy to the Roman republic, but the Egyptians kept aloof as much as possible from Rome; and while they avoided disputes, they equally renounced a participation in the honours of a close connection with the people who had overthrown the empire of the Pharoahs, and it is not till the third century of our era, that we find natives of Egypt accepting office under the empire. We may thus account for finding remains of Grecian civilisation in Cyrene, and of Egyptian architecture in Egypt; while all vestiges of Punic civilisation are lost, both on the African continent and in the Spanish peninsula. Turning our attention next to Egypt, we find in the remains of Antinoë very curious and interesting examples of Roman art. At Alexandria many fragments are constantly being brought to light, and used in works now in progress; and Pompey's pillar has been made familiar to us, both by pen and pencil. Abd-al-Latif, an Arabian physician of the time of Saladin, says, in his description of Egypt, that he had himself seen on the coast more than four hundred columns broken in two or three pieces, of which the material was the same as this column, and which appear to have been from onefourth to one-third the size. He adds that he could see by the fragments, that they had originally been covered with a roof. The translator states that this explains the origin of the Arabic name for the column-Amoud Alsawari, or the pillar of the colonnade; proving that it was originally placed in the centre of a forum, in the manner of that of Trajan at Rome. Denon gives a sketch of a column at the ancient Oxyrinchus, which appears of a colossal proportion, and with the remains of the architrave still existing on the capital; but the chapter in which we are led to expect the description gives us no dimensions.

The remains of Petra in Idumæa have been familiarised to us by the labours of Roberts and Laborde; but apart from their value as examples of a peculiar style, it is a subject of much interest to have a satisfactory elucidation of their date and of their history, in obtaining which careful search for inscriptions on the spot might no doubt materially assist. It is to Palestine, however, that we must look for a rich harvest of Roman architecture, in a field as yet almost untrodden by architects. Baalbec and Palmyra, the most important of its cities, have been to a certain extent investigated; but Palestine abounds with other Roman remains, which have been hardly sketched, much less measured and correctly delineated. At Antioch we find triumphal arches. At Missema, the remains of a small but beautiful hexastyle Doric temple, the interior decorated with four Corinthian columns. At Ezra, the ancient Zarava, the ruins occupy a space of three or four miles in circumference; among others a large quadrangular edifice with thirteen rows of arches, five in each row; and in every part of the town Greek inscriptions. At Amyouan, between Beyrout and Tripoli, is a tetrastyle Ionic temple, adorned with rich sculptures. Gerash appears to be one of the cities most fertile in architectural remains in this district, next to Baalbec and Palmyra. Among other objects is a temple near the gate, and facing it a large semicircular colonnade of the Ionic order, most of which, with the entablature, is still standing; the centre of this, exactly opposite the portico of the temple, opens upon the principal street of the city, also flanked by colonnades, and above a mile in length. There is also a large peripteral temple of the Corinthian order, surrounded by a double colonnade of smaller columns, in the manner of the temple of Venus at Rome.

At Damascus there is a fragment which has been already brought before the notice of the Institute. There are many others of which professional descriptions are wanting. The history of Baalbec and Palmyra is involved in much obscurity; and yet, architecturally speaking, these cities, with Petra, are among the most wonderful and interesting in the world. Petra for its extraordinary situation and character; Baalbec for the beauty of its style, and Palmyra for the unequalled extent of its remains.

In Asia Minor, notwithstanding the labours of recent travellers, what remains to be done in the investigation of ancient remains far exceeds what has been aiready accomplished. Some notice of the remains of Termessus have been already brought before the Institute; but in Caramania there exist some valuable remains of other cities; among others, those of Side. The walls are in some places perfect, and offer a curious example of ancient fortification, besides the usual accompaniments of an ancient city. There are

some interesting antiquities at Cacamo. A bath with piers supporting a vaulted roof of considerable space; and a granary built by Adrian or Trajan. It appears to have been customary to erect public granaries along the lines of the main roads for the supply of the troops on their march.

It is needless to dwell on the Roman works at Constantinoplethe aqueducts and cisterns have been often sketched but never measured. In Romania and other parts of the North of Turkey in Europe, as well as in Dalmatia and Istria, and wherever the Roman sway extended, monuments of more or less interest are to be found, the number of which might no doubt be increased by further investigation and research.

Remarks.-The CHAIRMAN observed, that Mr. Bell had opened a wide and interesting field, and many present had no doubt traversed some of the ground he had been over, and might be able to add some further interest to his remarks. It occurred to him (Mr. Fowler) to mention as an instance, that the road near Caudebec, in Normandy, passes through the remains of a Roman theatre or amphitheatre, having circular arcades. He had not seen any

notice of these remains.

Mr. TITE, Fellow-explained that they are situated at Lillebonne, anciently Juliobona, and are intersected by the old road from Havre to Rouen. There is in the same town a remarkably fine church of the Decorated period, of which he could not find a view or plan in any of the illustrated works on France. The members were much indebted to Mr. Bell for the pains and research exhibited in the paper just read. Mr. Tite had no doubt that much might be learnt of ancient art, out of Rome; but at the present day, we learn nothing of Roman architecture, either in Rome or out of it. Nothing but medieval architecture seems now to be the fashion, a circumstance which must be a matter of regret to all who have studied in earlier days a style which he considered infinitely better adapted to modern times and purposes. He would venture to say how necessary and essential he held it to be, that a young architect should study the remains of Greece and Rome. Mediæval art would no doubt afford useful principles of design and construction, but he could not conceive that a good architect conld regard his studies as complete, without a distinct investigation of the principles of Greek and Roman art. We are too forgetful of those principles in the present day, and therefore he the more valued the efforts made by the author of the paper just read; particularly as showing how much may be learnt in connection with Roman architecture out of Rome itself. With regard to Roman architecture in Spain, a work written in Spanish, by Pons, may be considered to contain an excellent account of the Roman remains in that country. It was printed in eight or nine small duodecimo volumes, about the and of the last century. Mr. Tite then moved a vote of thanks to Mr. Bell.

Mr. DONALDSON, Hon. Sec. For. Corr., observed, in reference to the introductory part of Mr. Bell's paper, that we should not supthat the Romans were employed solely in conquest, and not pose in diffusing a knowledge of the arts, because it must be well known that a great civilising spirit existed in them, and that wherever their conquests extended, they endeavoured to introduce good forms of government and municipal institutions. They expended large sums in the erection of monumental edifices in their provinces, even in England itself; and it must be a matter of regret that we have not a work worthy of being cited as 'Anglia Romana,' possessing as we do, a great number of buildings worthy to be recorded. If we had such a work, well illustrated, in the style which our works on mediæval art display, we should bring to light a number of interesting facts. Indeed, it ought to be a national work, promoted by the government. With respect to the monuments at Petra, he thought, from looking at the engravings that have appeared, that they present no evidence of Greek art at all, and that they must have been erected during the Roman dominion, in the time perhaps of the Antonines. With respect to the cities on the coasts of Asia Minor, reference may be made to the letters of Pliny, which recite the great number of buildings erected there under the Roman empire. It had occurred to him that all the buildings erected by the Romans out of Italy, are of a much lower class of art than those in Italy itself, and they are apparently of a later period, and have not the refinement of the Italian specimens.

Mr. SCOLES, Hon. Sec., mentioned that it had been ascertained that the column of Pompey at Alexandria, had been originally an obelisk, which the Romans had rounded and converted into a column; this became apparent on making some excavations underneath, when the hieroglyphics were discovered.

Mr. TITE, Fellow, said that Mr. Barry had made during his

travels a correct plan and a collection of sketches of the most accurate kind, of the ruins of Gerash and other neighbouring cities, which he had closely investigated.

Mr. G. GODWIN, Fellow, called attention to the excavations in progress at the Roman castrum, at Lymne in Kent, where, for want of funds, a Pompeii close at home is entombed, which might be opened for the satisfaction and instruction of all England

Mr. C. H. SMITH, Visitor, expressed his doubts as to the accuracy of the details and ornament given in the large works, under the names of Dawkins and Wood, though the measurements might possibly be correct. He had seen the original sketches some years ago, and observed that many of them were very slight, and that the drawings taken from them for the engraver, were made up in the style of ornament then in vogue, rather than in the spirit of the originals.

Mr. SCOLES, Hon. Sec., observed, when he was at Baalbec he had not that work with him, but his impression of the originals was, that they were equal, if not superior, to the finest specimens he had seen in Rome; the ornaments were elaborately and finely executed, and the Corinthian porticoes in his opinion where the finest in the world.

The CHAIRMAN, in announcing the vote of thanks, expressed his regret that Mr. Bell had not divided his ample and interesting subject into two papers, instead of condensing ít, in order to bring it within the scope of one evening's proceedings. He then made some remarks in allusion to the Palace of Diocletian at Spalatro, which, judging from the remains as illustrated in the work published by Adam, had evidently been erected after the decline of art in the Roman Empire, and which, though designated as a palace, was in point of fact, intended for a fortress.

NATIONAL PROVINCIAL BANK OF ENGLAND,

DARLINGTON.

THOSE Who look back to the last century will find, that whatever opportunities it afforded the architect for great monuments, it was far from yielding the same scope in street architecture as now. Churches, chapels, and mansions there were then, as now; but we are much better off with the banks, clubs, assurance offices, colleges and schools, not to speak of county courts, and many more classes of public buildings.

The subject we now give is not one of the more ambitious of its class, but is a very good example of the application of moderate resources. It is the building occupied by the Branch of that large company, the National Provincial Bank of England, in the respectable town of Darlington, and in which there are now two joint-stock banks. It is situated on the High Row, in a very conspicuous part of the town, facing the principal approach from the York, Newcastle, and Berwick Railway Station.

ton's disposal small; but by careful disposition and study, he has produced a building which, without pretension, is effective, and which it is none the less pleasing is completed within the estimate. We are very fond of columns when properly applied, but we are much better pleased in a composition of this kind to see that their employment is not attempted. It too often happens that stereotyped columns and pilasters are stuck on, by their ostentation to hide the architect's poverty of labour and resource; whereas, when such adventitious aid is rejected, there is always the hope of careful treatment. This, we consider, has been the result in Mr. Middleton's case, as the Elevation will show, and the details of some of which we have given engravings.

Section of the Top Cornice of the Front.

The management of the ground-floor is very good, and by attending to the breakings of the joints, the line of composition is carried up to the first-floor windows.

The treatment of the cornice, without being expensive, is rich; and the boldness of the proportion affords shadow and relief. The finish of the middle range of windows is likewise in good keeping. Whether the masques on the ground-floor keystones might not have been supplanted by emblems more significative, we leave to the architect to settle. Some local or commercial association might have had its meaning expressed.

Elevation of the Top Cornice of the Front.

The design for this structure was entrusted to Mr. J. Middleton, an architect practising in Darlington; and we are happy to have the opportunity of giving this proof of his successful application. The site is, it will be seen, narrow, and the means at Mr. Middle

Section and Elevation of Cornice to the First-floor Windows.

The building is of stone from the neighbourhood, and was erected this year. The internal arrangements afford the usual accommodations of a banking establishment, and for the domestic requirements of the resident and manager, Mr. M'Lachlan.

DEVELOPMENT OF GEOMETRICAL TRACERY.

On the Development of Geometrical Tracery. By the Rev. G. A. POOLE. (Paper read at a meeting of the Architectural Society of the Archdeaconry of Northampton)

Ir is sometimes objected to one who complains of a defect in any system of which he is treating, that he ought to produce a remedy for this defect. This, as a general proposition, would be at once rejected by every one, and yet, perhaps, every one is alike ready to apply it to those whom he does not affect, or of whose treatment of a subject he does not approve. And this, at least, must be admitted, that one who professes that he has seen the evil is not the last from whom the remedy may be expected. And having again and again felt, and professed to have felt, the inconvenience of the arrangement of Rickman, and of every architectural classification, where the style which intervenes between the Early English and the fully-developed Decorated is concerned, I shall now endeavour to justify my complaints, and to prove that there is such a generic difference between that style and the Early English and Decorated, on either hand, that it ought to have a distinct place in an architectural system and a distinct name in architectural nomenclature.

It is at once apparent that the styles of Gothic architecture are arranged very much with reference to the character of the windows. Right or wrong this is the case; and right it certainly is in the sense of being obvious and convenient; though it might perhaps have been expected that some more organic part of the structure might have afforded the characteristics of style. It should be considered, however, that the divisions of Gothic architecture are but sub-sections, or species; not kingdoms or genera. They are not analogous with the divisions of animals into vertebrate and molluscous, for this is parallel with the primary division of architecture into that of the arch and of the entablature. These grand divisions, then, being based on organic differences, it does not seem incongruous that the minor features of a building-even, if necessary, features far inferior in use and in powers of expression, to the window-should afford the differentials of genera and species.

The great point is, that the differences be constant and tangible; but here is the difficulty. There are facilities and difficulties in all systems, and in all parts of systems. It is easy to separate, in general, between a plant and an animal; it is easy to define the difference between the architecture of the arch and of the entablature; but there is a debateable province in both cases; in architecture the whole class of Romanesque buildings; in Zoology the countless species of zoophytes. Again, it may be as easy to distinguish, in general, between Decorated and Perpendicular as between a beast and a bird; but the buildings are countless which have as many of the characters of each style as the ornithorynchus has of the mole and of the duck. I wish this to be distinctly borne in mind all along, lest I should seem to fail in establishing a distinction; whereas, it is the very condition of all such distinctions that they shall have their vanishing point, not to the eye only, as where the sky seems to meet the earth at the horizon, but in the very nature of things.

And now what do we see, if we follow the forms of windows during the last half of the thirteenth and the first half of the fourteenth century? We see them gradually deserting the narrowness and simplicity of the lancet form, till, at last, they have arrived at a great variety and complexity, involving proportionate width of opening and the subordination of many parts. We see, in a word, a wide opening filled with mullions and tracery. And this tracery is composed, at first, of geometrical figures, following certain laws, and afterwards of figures no longer geometrical, and, though not without law, yet of that free flowing contour, which looks at least without restraint. Now, I think you will agree with me, that the first change and the last-the change from Early English to Geometrical, and the change from Geometrical to Flowing Decorated-both demand to be treated as the differentials of a style; the first, that is the mere introduction of tracery, as being, so far as windows are concerned, more important than the difference between Norman and Early English; the latter, the change of the laws which govern the formation of tracery, as being at least as important as any difference which separates Perpendicular from Decorated. In other words, Geometrical is more unlike Early English than Early English is unlike Norman; and so, ex abundanti, Geometrical and Early English should be separated; and, again, Geometrical is as unlike Flowing Decorated as Flowing Decorated is unlike Perpendicular; and, therefore, if

the two latter should be distinguished, so also should the two former.

And yet, the Geometrical is almost always treated as transitional (which, indeed, every style but the first and last must be, in some sense; but I mean that this is so treated as transitional, as if it had no claim to a name and station of its own); it gets no better title than Late Early English, or Early Decorated, as the case may be; the term Geometrical being only adjected to the generic term Decorated, as marking, not a genus, but a variety. If this had no practical result, it would be little worth contending about; but I believe that it really does result in the too great neglect of this style, as a model, and, at the least, a point of departure for modern practice. A style which deserves, but does not obtain, a substantive position, is sure to be defrauded of more substantial proofs of the estimation in which it ought to be held.

It is not my intention to enter at length on the process by which tracery was gradually evolved from the juxtaposition and grouping of several lancets. This has been done often enough. I assume that you are all well acquainted with it, and commence from the time at which Tracery, properly so called, was freely used; from the time, that is, when the portions of wall which separated lancets were attenuated and moulded into mullions, and when the piercings of window heads had left no portion of intervening stone-work of greater breadth than the interlacing of two equal tracery bars required.

And now imagine yourselves walking round some great minster at night, when the interior is lighted. I know no better way of coming at the effect of the windows taken apart from the rest of the fabric. Let the nave and south transept be Early English, but let the choir have been built towards the end of the 13th century, and, consequently, with windows filled with Geometrical tracery. As you turn the corner of the transept and get the first glimpse of the Geometrical choir, you feel yourself carried into a new age of design and of construction. But the north transept is Flowing Decorated, or Perpendicular, I care not which. As you leave the choir, and get a sight of this portion, there are differences, indeed, plain enough, even though the windows only are visible, but they are as nothing compared with the difference between the nave and the choir. Or, in other words, the difference between two kinds of tracery is as nothing compared with the difference between tracery and no tracery.

But, say some, the only appreciable differences are those of the windows. First, for argument sake, I grant it; but I have shown why differences in the windows may very well become differentials of style. But, secondly, in truth, I deny it. I deny that there are no differences of characteristic details between the Late Early English and the Geometrical, and between the Geometrical and the Early Decorated. And I deny this the more emphatically, because I shall not now stay to point out the differences: I shall merely ask you to take my word for it, that they run through every part of the structure, in composition, in detail, in decoration, even in construction-the latter, indeed, being demanded by the change in a matter of so great mechanical importance as the relative proportions of the windows, which, you will remember, are arched piercings of the outer walls, of no small relative magnitude.

But, at present, I confine myself wholly to the windows, and even yet more exclusively to the tracery, omitting even to notice cusping, the natural correlative of tracery, except where it follows the same laws as the tracery, which, in the Geometrical style, and in that alone, it often does-so much so indeed that a drawing of the tracery of one window may be converted into that of the cusping of another, only by altering the scale.

The first impression conveyed by a Geometrical window and a Flowing Decorated window side by side, is, that while the former is obviously drawn wholly with the compasses, the latter seems at least to be drawn in some degree libera manu. Perhaps this impression, so far as the Flowing Decorated is concerned, is hardly correct; but you will presently see that it results from certain appreciable causes, and indicates a real difference of principle in design. Take the simple Geometrical and an equally simple flowing two-light window. The eye at once detects the use of the compasses in the one, and the very centres from which the curves are struck; in the other no single curve is sufficiently simple to be referred, except with considerable effort, to its centre or centres; it seems, indeed, to be drawn without any mechanical aid. Take more complex arrangement, and still the same character is found carried out through 3, 4, 5, 6, 7, 8, 9 lights. This alone, as it seems to me, is sufficient to demand a separation of the two styles; for in speaking of design, this very fact, that the designer is put into so different an attitude as that of one who is limited wholly