Читать книгу Memorials of Old Lincolnshire - Various - Страница 28

Table of Contents

By A. Hamilton Thompson, M.A.

Lincolnshire is more rich than any other English county in churches which, if few are indisputably of a date earlier than the Norman conquest, retain traces of an architecture whose character at any rate is of a distinctly pre-Norman type. The county has nothing to show, it is true, of that early work, associated with the first century of Saxon and Anglian Christianity, which gives so unique an interest to the church architecture of certain districts in Kent and Northumbria. The neighbouring shire of Northampton possesses in Brixworth a monument whose importance overshadows that of Stow, and in Barnack and Earl’s Barton buildings which are, in point of detail, a match for Barton-on-Humber; while many who might hesitate to grant the pre-Conquest origin of Bracebridge, could hardly deny it to the Northamptonshire church of Wittering. The missionary visit of St. Paulinus has left but one trace, in the dedication to St. “Paul” of a church in Lincoln, of the connection of Lindsey with the religious life of Northumbria. That distinction on which Lincolnshire has prided itself, the possession of a Saxon cathedral at Stow, not hidden away in the foundations of a later building, but still in use as a parish church, begins to lose its value as the historical evidence on which it depends is more carefully examined. With the exception of Stow and St. Peter’s at Barton, the Saxon monuments of Lincolnshire are humble and unpretentious in character, without any very definitely architectural features; and he would be a bold man who should assert positively, on the little evidence which we have to show for their date, that they were built in days of Saxon rule. Equally bold was the assertion of Professor Freeman with regard to the two towers in the southern suburb of Lincoln, that native workmen built them while the Minster was rising in the style of the conquerors on the hill above.[30] Such positiveness is rebuked by the discovery that this picturesque statement was founded on evidence referring to churches in a different part of the city.[31] Avoiding either extreme, we may say, in our present state of knowledge, that most of the so-called Saxon churches of Lincolnshire represent a late state of Saxon art, open to Norman influence, but preserving a distinctly national tradition. Some of these monuments are undoubtedly later than the Conquest: of others, and perhaps of the majority, it may be said that, though a post-Conquest date is possible, yet the character of the work is of a kind that might be expected rather before the Conquest than after, and belongs at any rate to a type of art prior to the general spread of Norman influence. Thus the epithet “Saxon” may fairly be given to such buildings, even though, in point of date, they may belong to the Norman period.

St. Peter’s Church, Barton-on-Humber.

Of the pre-Norman date of the lowest stages of the tower and of the western forebuilding of St. Peter’s at Barton-on-Humber there can be no doubt, as we shall presently see. And at Barton the chief point of interest comes into view, in which these churches are of most importance to the architectural student. At the end of the tenth century, a date which may perhaps be claimed without extravagance for the Barton church, the parish church plan in England was a matter of experiment. The basilican plan with aisles had never been popular beyond a few larger churches: Lincolnshire does not furnish us with a single instance of the nave with aisles or of the apsidal sanctuary, which, with certain modifications, are features of some Saxon plans. The simple plan of aisleless nave and rectangular chancel, which had been adopted probably in most Saxon churches, had been complicated by the introduction of the tower into the scheme. The tower was for the present the uncertain factor whose place in the plan the Lincolnshire builders, and those of other counties with them, were trying to determine; and it is the position of the tower which gives Lincolnshire Saxon ground-plans their peculiar importance. We may be doubtful about the place of towers like Earl’s Barton and Barnack in the plan: they may have been merely western appendages to churches which have now been entirely replaced by later buildings, or they may have been the church itself, with its walls raised into a lofty tower, at once a place for bells and a look-out in time of danger. There can be no such doubt at Barton-on-Humber. In our own day the tower and its western annexe stand at the west end of a large late Gothic church; but, at a restoration in 1898, the foundations of an eastern annexe, very similar in size and shape to the western, were discovered, showing beyond all doubt that the tower formed the main body of the church, standing between a chancel and what may have been a baptistery—rooms for the altar and the font—of almost equal size. The tower was broader than its adjunct: two entrances remain, one on the north, the other on the south, opposite each other, and towards the western end of either wall. Mr. Hodgson Fowler, who discovered the foundations of the chancel, also discovered other foundations, presumably of Saxon date, to the east of the tower, which seem to suggest that a larger building with an elongated plan was in contemplation, but was superseded by a compact plan which found itself centralised in the space allotted to the tower.[32]

A somewhat similar plan occurs at Broughton, a village some four miles west of Brigg, and close to the line of Roman road which led from Lincoln to the Humber. Here the fabric has none of those distinctively Saxon features which are found at Barton; it is almost certainly a work of later date. The eastern wall of the tower, as at Barton, has been absorbed in the breadth of an aisled nave of a later period; but the eastern quoins are still visible, continuing to the ground, and indicating that the tower was once broader than the portion of the building east of it. That this eastern building, moreover, was small, and that the bulk of the congregation occupied the space west of it, is suggested by the fact that, as in many later chancel arches, decorative treatment—too rude here to be taken very seriously as architectural membering, in spite of its efforts—is confined to the western face of the arch by which the chancel was approached from the tower-space. Here, then, we have once more the tower-space forming the main area for worship, with a small chancel to the east. At Broughton, however, instead of the corresponding western annexe which we find at Barton, there is a large three-quarter-circular projection, containing a newel staircase which leads to the belfry stage of the tower; and, instead of the two doorways at Barton, there is only one, this time in the south wall. The western projection at Broughton is often ridiculously called a western apse. It is, and always was, a turret for a staircase. Three other such turrets exist. One, at Brixworth in Northamptonshire, was added in front of an earlier western doorway, when the original porch of the basilican church was heightened into a tower. Another, at Brigstock in the same county, forms an integral part of a western tower with strongly Saxon features; its staircase was always of wood, as may have been originally the case at Broughton.[33] The third instance is in Lincolnshire, at Hough-on-the-Hill, seven or eight miles north of Grantham. Here the stair turret is part of a western tower of more than probably pre-Conquest date, and of proportions as ample as those of Earl’s Barton; and one is tempted to discover a parallel to Broughton, and another quasi-parallel to Barton-on-Humber. But the face which the tower presents to the church behind it is singularly blank; and it remains to be seen whether there lie hid, beneath the plaster, quoins, like those at Broughton, to indicate the existence of a chancel whose foundations may still be buried beneath the western floor of the nave.

At Barton and Broughton, and possibly at Hough, we are face to face with a small compact plan—at Barton definitely centralised in the tower-space, at Broughton without the same centralisation, but with the main body of the church still gathered beneath the tower. In these cases, when we speak of the tower and the tower-space, we must regard the tower simply as an upward continuation of the body of the church. The congregation has not found shelter on the ground-floor of a tower: the tower is the upper storey or storeys of their church. However, in the two further instances of towers not western which Lincolnshire affords us, the tower-space must probably be regarded as a feature in the plan distinct from the main body of the building; it is not a church on which a tower has risen, but a space which is there because a tower forms a definite part of the design. Waith Church, a few miles south of Grimsby, is for the most part a modern Gothic building, with an entirely modern plan.[34] But between the chancel and nave, and flanked by a south transept, rises a tower of a type very familiar to travellers in Lincolnshire, but here alone seen in a central position. Its position must always, however, have been between eastern and western out-buildings; for its eastern and western walls are pierced by low arches of equal height and width, very different in proportion from the ordinary western doorway and tower arch. There seems to have been no entrance in either of the side walls. In all probability, then, we have here, not a definitely centralised plan, but a tower-space intervening between a nave and chancel. Of the relative dimensions of these to the tower-space it is impossible to speak: we have no remains to guide us. Again, this tower may have been simply an elevation of the eastern portions of the nave walls, as in those cases to which the term “axial” has been given; or, as in some Norman churches, it may have projected north and south of the adjacent nave and chancel. In the last case, we should have the ground-plan of Barton-on-Humber, with its centralised character probably destroyed by the elongation of its western annexe. The nave and the tower-space become independent divisions of the plan.

St. Mary’s Church, Stow, looking East.

The Saxon church of Stow survives only in part; and to assert that the present fabric, which is largely of the later part of the eleventh and the earlier part of the twelfth century, is a rebuilding of the older church on its original scale, would be to assert what we do not know. However, the church was planned on a scale somewhat more imposing than was usual in Saxon times; and enough of the older work is left in the transepts to assure us that they, at any rate, covered their present site from the date of its foundation. Their length and general proportions postulate a nave to match; and we may assume, without much doubt, that the present Norman nave rose upon Saxon foundations. The chancel may have been enlarged to its present dimensions by Norman builders; this is, at least, more likely than that the Saxon chancel was equally spacious. The visitor to Stow about the time of the Conquest would have seen nave, chancel, and transepts, as indeed the visitor to-day may see them, grouped round a central tower, which rose straight from the ground in their midst, independent of their buttressing aid. The quoins of the tower go down to the ground; the arches which connect the tower-space and the adjacent arms of the building are, as it were, piercings in the tower walls rather than the actual substructure on which the tower walls rested.[35] The tower-space at Stow is thus in some measure a central area, the focus of the plan; and a vivid imagination might conjure up in this instance the Barton-on-Humber plan reproduced on a larger scale, and converted into a Greek cross by the addition of transepts. But it is more probable that here, as at Waith, the tower-space is shifted slightly to one side of the centre of the plan, and, while keeping much of its dignity in the general scheme, is no longer the main body of the building.

In most English churches the most convenient plan from the earliest times has been the oblong nave and practically square chancel, divided by an arch which, to our modern ideas, has sometimes been inconveniently narrow, but without the intervening tower-space, which became in so many later churches an obstruction to the unity of worship in chancel and nave. We have seen Lincolnshire builders experimenting with that new-found addition to the plan, the tower, packing their nave into its ground-floor, trying what can be done with a central area, abandoning—we speak of probabilities—the complete symmetry of the centralised plan, and finally wedging the tower in between the arms of the building, as an effective focus for the church as seen in elevation. The difficulties, the inconvenience, the uncertain conditions, of centralised or quasi-centralised planning, are now in most cases abandoned: the builders frankly remove their tower-space to the west end of their plan. Upon it rises a bell-tower, which may on occasion be used as a look-out tower in time of disturbance, or even—though this seems very doubtful—as a place of refuge for the inhabitants of the township. In most instances the tower-space will be entered by a western doorway, and will be the porch of the church, just as, at Brixworth or at Monkwearmouth, in other counties, the original porch has become the substructure of the tower. The porch will lead into the nave of the church, oblong and aisleless; and, in the east wall of the nave, an arch will give access to a small rectangular chancel. This is the normal Lincolnshire, and indeed the normal English plan; and this plan powerfully affects the architecture of the Norman and Gothic periods of English art. The centralised plan may survive in beautiful forms, and will always be the more interesting, owing to its greater capacity for variation; but the western tower of the Saxon period, and the elongated plan associated with it, will be the standard of planning congenial to the larger number of English masons.

It is unnecessary to particularise between the various churches of Saxon origin in Lincolnshire which have western towers. There are many, and the number may be stated rather variously. The present writer, excluding Hough and Broughton, which, as we have seen, may be treated more suitably with centralised plans, counts some thirty towers in part or wholly of the distinctively Saxon type.[36] Some of these, as he already has said, evidently were built at a date later than the Conquest. Of no one of them would he courageously assert, on the mere evidence of plan and details, that it was built actually and beyond doubt before the Conquest. But that they were built by the hands of Saxon workmen, and that they represent a definitely Saxon tradition, are hypotheses which, if they do not offer themselves to a very clear proof, may at any rate be enunciated as highly probable.

The consideration of the dimensions of these towers on plan may be left to the discussion of their relative dates, with which this chapter will conclude. Having noted variations of plan, we must now look at architectural details. Of those peculiarities of technique which are most readily recognised as Saxon, St. Peter’s, at Barton-on-Humber, is a nearly unique example in Lincolnshire, and its value is still higher, in that the upper stage of the tower presents features of a rather different kind, more typical of Lincolnshire, but less specially and exclusively Saxon than those of the lower stages. The tower is divided by two string-courses into three stages, the middle stage low and squat, the lowest stage much the tallest of the three, and subdivided into two parts, an upper and lower, by external decorative arcading. This subdivided stage represents the body of the church; the middle stage probably represents the original bell-chamber; and both these stages, together with the small western annexe, have definite “long-and-short” quoining. The “short” stones, as usual, back into the rubble-work, of which the tower is built; but their protruding faces are cut away flush with the rubble, and are hidden beneath the plaster which covers the whole surface of the tower. The decorative arcading, however, already alluded to, is formed by irregular strips of dressed stone projecting from the surface, the heads of which, formed by small horizontal impost-blocks, are connected in the lower stage by semicircular strips. On the crown of each of these rude arches rests the foot of one of the upright strips of the upper stage, which are connected similarly by strips of triangular form, the apices of which touch the under side of the string-course between the lower and the middle stage. The surface of the lower stage is thus cut up into two series of tall arcaded panels. The bottom part of one of the lower panels is pierced on the north and south sides of the tower by a doorway with rounded head. The upright, dividing two of the upper panels on each of these sides, is partly cut away to make room for a double window-opening with rounded heads, the opening being divided by a small piece of wall faced, at the level of the outer wall, with a baluster-shaft. These windows lighted the body of the church, the inner roof of which came at this point. The middle stage keeps the “long-and-short” quoining, but the strip-work has here given place to an unpanelled plastered surface, broken only by a double window-opening, similar in construction to that in the stage below, but with triangular instead of semicircular heads. Like the middle stage of the tower, the western annexe of the church has no strip-work on its walls, but has “long-and-short” work at its angles. It is lighted by a semicircular-headed opening in each of the north and south walls, and in the west wall by two circular openings set one above the other. All these openings are splayed outwards as well as inwards. The eastern wall of the tower can be seen from the inside of the present church, with its “long-and-short” quoining perfect to the ground, and with breaks in the masonry where the eastern annexe originally joined it. The arch which pierces it on the ground-floor—the chancel arch of the Saxon church—is very plainly treated with dressed jambs, impost-blocks, and voussoirs, but without any moulding. In the wall above is a single opening of considerable width, with rounded head, rather massive jamb-stones, and thin, flat impost-blocks. Above this comes the double opening of the belfry stage, which would have stood clear of the roof of the Saxon chancel.[37]

Turning from these features of the original church walls, its western annexe, and its belfry stage, to the uppermost stage of the tower, we are met by a striking difference. We already have seen the strip-work of the lowest stage disappear. Here the “long-and-short” work is gone as well, and the quoining is of small oblong stones set on one another at right angles, so that each of the adjacent faces of the wall is in bond with every other of the quoins. The window-openings are still double, and have rounded heads, but they are taller than those below, and are divided, not by slabs of wall with baluster facings, but by slender rounded shafts set in the middle of the thickness of the wall, with heads corbelled out so as to form rude capitals, and to support through-stone impost-blocks, corresponding to those at the head of the jambs on either side. Of the absence of splay, inner or outer, to the openings we can say nothing; the double splay has occurred only in the western annexe. But the disappearance of “long-and-short” work, that most unmistakable of purely Saxon details, and the introduction of a new type of double opening, are significant of a change of style which has come over the Saxon building art since the church and tower began to rise.

Thus, at Barton-on-Humber, we have two different types of Saxon work—that very peculiar form, with its tendency to panel decoration with strip-framing, which produces its highest decorative effect at Earl’s Barton, side by side with a more staid, less fantastic manner of building, which is without architectural ambition, uses decoration very sparingly, but can achieve very pleasant effects of proportion within its modest limits. This second style, as it may be called, is emphatically the style favoured by Lincolnshire builders. Of the first style, Barton-on-Humber is the only really conspicuous example in the county.[38] Strip-work decorations, not uncommon in the Saxon work of the South Midlands and South of England, of Mercia and Wessex, is quite the exception within the belt of Danish influence. It appears here and there as a kind of frame to arches and their jambs, or to the heads of window-openings. The best examples of its use in this connection anywhere in England are to be found in the jambs of the noble tower arches at Stow, where a semicircular shaft is carried down the face of the wall close to the angle of the jambs, and is accompanied by a flat strip of stone at a few inches distance. Both shaft and strip are finished off by rough corbels a little above the floor level.[39] But Stow is an exceptional church. As a rule, we find the strip-frame retained purely as a flat hood-mould to doorways and windows, without a trace of that individuality of style which distinguishes it at Stow, and preserving a still more distant kinship to the work at Barton.

“Long-and-short” work pursues a more hardy existence. Quoining was necessary, and the “long-and-short” method was at once serviceable and fairly ornamental. So, while strip-work, a merely decorative arrangement of pilasters without constructive use, went its way, “long-and-short” quoining remained. We come across it chiefly at the angles of naves, which in several cases have been left almost untouched, when aisles of a later date have been added. St. Mary-le-Wigford and St. Peter-at-Gowts at Lincoln are cases in point. Bracebridge is an excellent example, for here all four angles of the nave can be traced. Cranwell, near Sleaford, and Ropsley, near Grantham, are other unmistakable instances. But here we must beware. The critic is too common who, assuming that a piece of wall is Saxon in character, immediately jumps to the conclusion that its quoining must be “long-and-short.” If the quoins are not arranged in a regular series of pieces alternately vertical and horizontal, then the work is not “long-and-short” work. If one or two stones thus arranged occur in the middle of irregular or of the common “small-stone” quoining, we are not justified in speaking of the fabric as showing “long-and-short” work. If the quoining shows a merely rough general resemblance to the “long-and-short” arrangement, it is not “long-and-short” work, but work of a quite haphazard type.

This brings us to one of the leading features which distinguish the towers so characteristic of Saxon work in Lincolnshire. We may study their angles to our heart’s content, and discover “long-and-short” work with the eye of faith, but we shall actually see it in only one instance, and there in the jambs of a western doorway of the tower, rather than in the quoining of the tower itself. In this instance, at Rothwell, near Caistor, the south-west quoining of the adjacent church has been left standing, like a small rectangular buttress, against the junction of the twelfth century nave and south aisle. It is formed of irregular stones, but such “long-and-short” work as there is, is confined to the tower. This is an exception. At the Lincoln churches and Bracebridge, where we have noticed “long-and-short” quoining at the nave angles, the quoining of the tower is of small stones; and this is universally the case. If we go northwards, by Marton, Heapham, Springthorpe, Corringham, Harpswell, and Glentworth, to Winterton and Alkborough; if we cross the Ancholme to Worlaby, and then go by Barton-on-Humber to Clee, Scartho, Holton-le-Clay, Waith, and Laceby; if we traverse the Wolds by Swallow, Cuxwold, Rothwell, and Cabourn to Caistor and Nettleton, and, descending by way of Hainton to Lincoln, make our way along the South Cliff to Branston, Harmston, and Coleby; if we go as far south as Boothby Pagnell, Little Bytham, and Thurlby-by-Bourn, and finish our journey in the midst of the parts of Holland at Great Hale, the only genuine piece of “long-and-short” work we shall have found in a western tower is that at Rothwell. It is true that this journey will have included more than one doubtful member of the family, and some of its genuine members which have lost, under the hand of the restorer, most of their appearance of age; but its result will be the establishment of the general rule that the Saxon tower-builder in Lincolnshire did not avail himself of the “long-and-short” method of quoining.

It will hardly need this journey to be convinced of his preference for the double window-opening, divided by the mid-wall shaft. This declares itself patently in the well-known towers at Lincoln; and all the towers mentioned above still have, or probably have had, such windows. Sometimes, as at Nettleton or Coleby, the belfry stage has been entirely renewed in the later Gothic period. Sometimes, as at Winterton or Alkborough, the tower has simply been heightened, and the Saxon belfry stage has become an intermediate storey. Sometimes, as at Cuxwold, the top of the tower has been lopped off altogether, or, as at Swallow, has been replaced by a modern stage in a rather incongruous style. In every case the original existence of the “mid-wall shaft” window cannot be reasonably doubted. The form of such openings as remain is very much the same. Its main outlines have been seen in the uppermost stage of the tower at Barton: two adjacent openings, with dressed jambs and voussoirs flush with the general surface of the wall, with rounded arches springing from through-stone impost-blocks, and received at their meeting by another such block, which rests on the mid-wall shaft itself. These openings pierce the wall without any splay. They have no strip-framing, and seldom, if ever, any attempt at hood-moulding. Although, as has been hinted, some beauty of form may be claimed for them, they are as simply constructed a type of arched opening as could well be devised. Their proportions are sometimes rather elegant; and, when they are set round a small upper stage, divided by a projecting string-course or off-set from the unbuttressed and sometimes slightly tapering length of the lower stage, their effect is always striking.

The architectural value of these towers, so simple in their principles of construction, so insignificant in their height, is less than their historical interest. Saxon builders had little architectural knowledge or skill; and buildings like Stow impress us more by their height and mass of wall than by any very striking architectural feature. The work at Barton-on-Humber is curious and interesting building: it is not architecture. In the Lincolnshire towers, a step is taken in the right direction by the avoidance of merely decorative surface-ornament. The tower asks for judgment on its own merits. Where it is divided by offsets into two or three stages, the result is satisfactory; although, if the belfry stage is of much the same area as the stage below, the tower looks top-heavy. This certainly is the case of St. Mary-le-Wigford. At St. Peter-at-Gowts, a small upper stage is set firmly and squarely upon a long and tapering lower one; and there is no finer tower in the whole series. The third type, where there is no off-set—the much-restored tower of Springthorpe is now, if it was not always, in this state—is merely insignificant.

The treatment of openings in these towers, other than the mid-wall shaft windows, is open to few variations. Western doorways are low and narrow: large stones are used in the jambs; and, though the heads are arched, the actual opening is covered by a flat lintel. The roughest of these openings is in the tower at Winterton, where the head of the doorway is formed by a huge stone, cut with a segmental curve on its under side to give the effect of an arch. Flat rectangular hood-mouldings of small projection sometimes follow the curve of the doorway arch and meet the extreme edge of the impost-block. At Clee, one of the best towers of the group, such a hood-moulding bounds a doorway head of two orders of voussoirs, the lower slightly recessed beneath the upper; but such refinements are rare. A similar recessing of a lower band of voussoirs occurs in the tower arches at Clee and Scartho, but in no case is it accompanied by any attempt at moulding the arch or recessing the jambs to match. An edge-roll was worked very tentatively at Nettleton along part of an unmoulded tower-arch, but was abandoned when about half completed. An ambitious and unique attempt at recessing, in the chancel arch at Broughton (now the arch from the tower into the nave), remains as a monument of the failure of the Saxon mason in his search for means of architectural expression. Both orders of the arch spring from an undivided impost-block, and the shafts, which should bear, and are intended to correspond to, the inner order, are stranded on either side of the back of the opening, with their heads left bare and their function denied them. As a rule, in doorways and tower arches, the mason was content with a plain unmoulded arch, springing from projecting impost-blocks on the top of jambs, the dressings of which are simply the quoins of a rubble wall. He varied the proportions of his tower arches, giving them great height, breadth, and dignity at St. Peter-at-Gowts, building them tall and narrow at Clee, Scartho, and Holton-le-Clay, or with rather less elevation and rather more breadth at Rothwell and Cabourn, frequently allowing them, as at Cuxwold or Alkborough, to remain low and rather broad in proportion. These variations of the tower arch constitute one of the most interesting features of this type of building: they introduce an element of individual design, and the loftier form of arch often produces by its mere size an effect which is not due to any obvious architectural virtue.

The lesser windows of the towers are usually small and narrow, with an inner splay. Their outer openings are often flanked by very large dressed stones: their rounded heads are seldom arched—there is a good arched window head in the south wall of the tower at Coleby—but are more often cut in the under side of a lintel; and sometimes this cut, exceeding a semicircle, produces the “key-hole” form of opening. At Rothwell there is in each wall of the tower, below the “mid-wall” windows, a small rectangular opening with a wide inner splay: a somewhat similar opening pierces the wall above the west door at Nettleton.

In the masonry of these towers two striking features are apparent. One is the disappearance of that “through-stone” treatment of dressed masonry, which is an undoubted characteristic of early Saxon work. A little doorway in the west wall of the north transept at Stow has voussoirs and jamb-stones, each of which faces the whole thickness of the wall. But the jambs below the tower arches are faced with double or triple, not single stones. And while it is rare to find a tower arch or doorway of this style formed of a core of rubble between facings of dressed stone, yet there are few in which the facing stones do not become less closely set together, and wide rubble fillings do not take the place of neat and close jointing. The other feature is the appearance of “herring-bone” masonry. This may be seen in some profusion at Broughton, and in a striking and unusual form at Marton; and, although it is not general, it occurs in other places.

Modifications in the tower plan are almost confined to an increase of dimensions which, in some members of the group—notably Caistor—is rather remarkable. One tower alone—Great Hale—introduces a newel staircase into an angle of the fabric; and this is almost absurdly unsuited to the probable size of those who had to climb it. We have seen that, at Hough and Broughton, nearly circular excrescences were formed to hold stairs on the west side of the towers. As a rule, we may believe that the upper floors of the towers were approached by ladders. It may be noted that in Lincolnshire there are very few of those openings above the tower arch, which are often quoted to prove the use of the tower as a place of habitation, and probably led to a landing and wooden stair communicating with the interior of the church. There are such, as at St. Peter-at-Gowts, Winterton, and Broughton; but they are exceptional, and the probable plan of Broughton makes it possible that, as at Barton-on-Humber, the opening was merely a piercing in the wall between nave and chancel.

But if there is little variation in plan, there is, as has been noted, even in the simplest towers, some degree of variation in detail. There are cases, moreover, in which the tendency to variation takes the direction of increased ornamental treatment. Instances are quite common in which the heads of the mid-wall shafts, bulging to support their impost-blocks, have been carved into the form of capitals—plain cubical or cushion-blocks, as at Winterton or Clee; rough suggestions of classical volutes and foliage, as at Glentworth or Scartho; varied forms of fruit and leafage, as at Bracebridge; or delicate and cleverly cut relief work, as at St. Peter-at-Gowts. And these capitals are not the only features which show the tendency. At Alkborough ornamental material has been transferred from some deserted Roman villa in the neighbourhood: a cornice from an entablature has been cut up, used as imposts, and turned upside down to form plinths for the jambs of the tower arch. But at Branston, south of Lincoln, the builders have not borrowed ornament. Like the masons at Broughton with their abortive recessed arch, they have tried to copy what they have heard of, what some of them at any rate have seen, and have covered the lower part of the west wall of their tower with arcaded panelling, with rounded arches and cushion capitals. This, and the central doorway with shafting and rudely foliated capitals, have been inspired from a source quite distinct from that which brought into being the strip-work panelling of Barton-on-Humber.

We are at once impelled to ask what this source is. And this question brings us to the consideration of two final questions, which are complementary to one another. What influences from the Christian architecture of other countries were felt by Lincolnshire masons? Is there any element of progress to be traced in the Saxon buildings of the county? In short, to combine the two questions into one, can any chronological sequence be traced in these buildings, by comparing them with the work of Romanesque builders in other countries?

We already have allowed the term “Saxon” to them, on the understanding that a pre-Conquest date is not implied thereby, but merely the fact that their style is different from that of buildings to which we give the term “Norman.” It may also be premised that, in considering their relative date, we have to deal cautiously and tentatively with a series of probabilities. We must also put legend aside. A hardy tradition, resting on no authentic basis, but engraved on a brass tablet within the church, points to certain traces of fire in the church of Stow as evidence of its burning by the Danes in 870. If this were true, and if the lower walls of the tower, as they exist to-day, were the tower walls of the Saxon cathedral of Sidnaceaster, we should be able to point to a church of the ninth century, if not earlier, which would probably have supplied an architectural standard to the diocese. However, we have nothing but the size of the church and an unfounded, if time-honoured, assumption to give it claim to cathedral rank. The very name of Sidnaceaster was probably invented in post-Conquest times, by some one who misread the signature of one of the bishops of Lindsey to the decrees of a Saxon council.[40] And, finally, the authentic history of the church of Stow does not begin till about the year 1040, when Eadnoth, Bishop of Dorchester, with the powerful assistance of Earl Leofric and his wife Godiva, set a band of religious men on the site hallowed by memories of the miraculous sojourn of St. Etheldreda on her flight from York to Ely. All that lies behind 1040 in connection with this church is pure romance. When Bishop Rémi later in the century restored St. Mary’s Minster at Stow, what he probably did was to complete Eadnoth’s ambitious beginnings, on a worthy scale, out of reverence for St. Etheldreda, and not out of sentimental feeling for an old cathedral church which he had superseded for all time by his new church at Lincoln. If a small archdeaconry in the Norman diocese of Lincoln, corresponding to the original district of Lindsey, took its name from Stow, we must not consider this as admitting the old cathedral dignity of Stow. The size of Eadnoth’s and Rémi’s church made it the most prominent building in the archdeaconry: no more convenient or suitable name could be supplied to the district.[41]

1040, then, is a recognised date to which we can refer the earliest work at Stow. We know, too, that a church was founded at Alkborough in 1052, and the work in the western tower there, with its triangular-headed belfry windows, may be claimed for that date or not long after. These are practically the only two pieces of dated evidence on which we can rely, for we have already seen that the evidence as to Coleswegen’s churches at Lincoln does not apply to existing buildings.[42] Stow, as we have seen, retains strip-framing to the tower jambs on a large scale: “long-and-short” work occurs in the north transept doorway and in the jambs of a window opening in the south transept: through-stone masonry is used in the north transept doorway, but abandoned in the tower arch jambs. Alkborough has none of these characteristics. We are thus at liberty to assume a gradual cessation of purely Saxon technique between 1040 and 1052.

Further, at Barton-on-Humber, although we have no documentary evidence to guide us, it is obvious that the tower is of two styles. The uppermost stage forms an addition to the original design in a somewhat more simple style. An interval of date between the stages is certain: the length of that interval would be hard to ascertain. But the fantastic strip-panelling and “long-and-short” work of the lower stages of the tower belong to the height of a fashion in architecture which is seen gradually disappearing at Stow. The strip-framing and “long-and-short” work at Stow are of a different and less purely decorative type: we are not surprised when in the upper stage of the tower at Barton, or in the tower at Alkborough, they disappear altogether. So far, then, this statement of progress is justifiable. The lower stages of the tower at Barton are obviously earlier than the upper stage. The upper stage has affinities of detail with the tower at Alkborough, and clearly belongs to much the same or a slightly later period. The tower at Alkborough is at least twelve years later than the only trustworthy date for the early work at Stow. And, as we have just seen, the work at Stow, in its selection and treatment of elements which had been used at Barton in careless profusion, is probably later in date than the earlier portion of Barton. With Barton may be grouped, from the character of its “long-and-short” quoining and the window-openings of its stair turret, the interesting tower at Hough-on-the-Hill. With Stow we may group, at any rate provisionally, those fabrics which have “long-and-short” quoining of a substantial type, flush with the surface of the wall, instead of projecting in rather thin strips beyond it—this will include, as we have noticed, some naves of churches. With Alkborough and the upper stage at Barton can be combined church towers generally, Hough alone excepted, and work of a partly Saxon character, like that at Stragglethorpe, not far from Hough. The “long-and-short” work in Rothwell tower is so small in quantity that it can hardly be treated as an exception to the third group.

Professor Baldwin Brown, in his valuable monograph on Saxon ecclesiastical architecture, has provided strong arguments for the influence of Teutonic Romanesque architecture on our Saxon builders.[43] It has long been the fashion to suppose that the decorative detail at Barton-on-Humber and other kindred churches is an imitation of timber construction in stone; the rudeness of treatment makes the supposition excusable. But we can hardly grant that Saxon builders could have imitated a system of construction which was not at any rate general till a much later date; and it is much more likely that the work at Barton is copied roughly and clumsily from a type of decorative work which, in the tenth and early eleventh centuries, was common in the Rhenish provinces and in the districts of Northern Italy architecturally related to them. It is unquestionable that the double opening with mid-wall shaft found its origin in the same provinces. Northern Italy doubtless exercised its influence on Germany; Germany, in turn, influenced the Saxon masons. In three other cases at least German influence is more than probable. It is certainly responsible for the double-splayed window-opening; it probably affected the simple type of capital which was developed into the cushion capital of later days; the unfaced rubble masonry and thin walling of Saxon churches are found commonly in early Romanesque German churches, and are the antithesis of the faced rubble and thick walling of Normandy and the Romanesque buildings of France; while the position of the western tower in the Saxon plan is a further German feature.

On the details of English intercourse with Germany it is unnecessary to dwell. There is plenty of evidence to show that a close connection existed between the two countries, which cannot but have had influence on the progress of art in England. That progress must have been practically at a standstill during the long epoch of Danish invasion. The date at which we may most reasonably expect an architectural revival to be general is that period, the last thirty to forty years of the tenth century, when so many monasteries were rebuilt or newly founded, and the monastic life re-established. This revival was strongly influenced by the monasteries of the Netherlands, which lay in the direct current of architectural progress between Germany proper and England. To this date at earliest, then, may be assigned the earliest parts of the church at Barton-on-Humber, and possibly the tower at Hough. We cannot go further back without deserting probability. At the same time, if we limit this work to the latter half of the tenth century on the one hand, we are not precluded from allowing that it may be later. Barton-on-Humber lay in the very path of the Danish invaders who established their power in England between 1002 and 1014. The base of operations of the heathen Swegen was at Gainsborough; there he died, and there he was said, though wrongly, to have been buried. The older parts of the church at Barton show no sign of the ruin which we might expect to have thus befallen them. Perhaps the church was left roofless by Swegen, but restored and completed, with the upper stage added to the tower, towards the middle or after the middle of the eleventh century. But it is also equally probable that the whole lower structure may be a rebuilding under the Christian Canute of a church ruined by his father; and that, after an intermediate stage in which the church may have taken the form shown by Professor Baldwin Brown, the tower was heightened by a storey.

The oldest Saxon fabrics in Lincolnshire need not, therefore, be earlier than the eleventh century. It will be noticed that of these Barton gives us a centralised plan, while Hough presents a plan which certainly is not in keeping with that of the usual western tower. The head of the second group, Stow, is another experiment in “central” planning. The remaining anomalies of plan, Broughton and Waith, belong, in the matter of technique, to the third group, in which the western tower predominates overwhelmingly. The German features of the third group already have been described. But they are less marked than those of earlier groups, and the buildings are open to influences, especially decorative influences, of quite another kind.

Towers like those of St. Mary-le-Wigford and St. Peter-at-Gowts, or of Bracebridge, may be regarded, on the strength of their different quoining and their lack of any bond with the fabric behind, as final additions to churches which, in point of date, we have provisionally classified with Stow. Not infrequently, the tower and church were built together without afterthought, as at Winterton, where the ends of the nave walls still remain in bond with the tower, enclosed within the spacious church of the twelfth, thirteenth, and fourteenth centuries. If we attempt to classify these towers chronologically, however, we may try several standards. The relative height and width of the tower arch affords no criterion; this seems to be varied at the fancy of local masons. Again, rude attempts at moulding, like that at Nettleton, tell us nothing; the moulded tower arch at Corringham is so out of keeping with the tower itself that it can hardly be part of the original design; the wide arch at Harmston is clearly a reconstruction achieved comparatively late in the twelfth century. But the absence of through-stone masonry, the approach to the system of rubble core and dressed facing, indicate a growing assimilation to “Norman” methods. The tower at Waith has a facing of rough ashlar, which is found in no other Lincolnshire tower of the type. Another tell-tale sign is the appearance of herring-bone work, which Norman builders used freely in the walling of their earlier castles, and builders of the pre-Conquest period certainly used little, if at all. The “herring-bone” masonry of the tower at Marton is identical in style with that in the curtain-wall of Tamworth Castle.

Can these novelties have come into use before the Conquest? If so, we must presuppose a growing acquaintance with building methods as pursued in Normandy. This is not improbable. Norman influence was felt in English life during the reign of Edward the Confessor,[44] and imitations of Norman, as at an earlier date of German, technique are not beyond reasonable imagination. But it must be acknowledged that, while we are fain to discover the germ of Anglo-Norman art in these simple monuments, the constructive, purely architectural quality, which is the essence of that art as manifested at Durham or in the west front of Lincoln, is totally absent from them. That essential quality is not a home growth, but a foreign importation. The chief element of structural transition in these towers is to be found in their growing spaciousness. Areas which are on an average oblongs of some 10 feet 5 inches long by 10 feet 8 inches broad grow at Caistor to 15½ by 17½ feet, and at Harpswell to about 15 by 16 feet. With the growth of the area the wall thickens, thus affording a contrast to towers like Hough, where the large area is enclosed by thin walls. The eastern wall of the tower at Alkborough is less than 2½ feet thick; the average thickness of such walls is about 3 feet 7 inches. At Caistor the wall is 5¾ feet thick. One cannot fail to recognise that the slender tower of Saxon type, while keeping its unbuttressed character and its “mid-wall” arches, is spreading out into the ordinary broad and heavy Norman tower.

In the capitals of the mid-wall shafts a decided step is being taken in the same direction. The plain “cubical” type may be regarded as following a course of natural development at home which found conspicuous perfection in the architecture of the Norman period. But in other examples, and those fairly numerous, the end aimed at is clearly that reminiscence of the Corinthian capital which is familiar to those who have visited the abbey churches at Caen, with its upper volutes and its lower band of acanthus. The best examples of this are at Scartho and at St. Peter-at-Gowts, where the effort, if somewhat crude, is still to some extent achieved. The effort is again seen in the capitals at Great Hale, in which the general outline is that of the “cubical” cap, with volutes carved on the flat upper part, and the carved under surface reeded so as to give a suggestion of the band of acanthus. At Bracebridge, again, in capitals of the “cubical” type, volutes and rather unusual forms of conventional foliage appear; and in other cases, as at Glentworth, the volute is used without much skill, but with some variety of design. The unfinished form in which the sculpture is sometimes left suggests forcibly that it was added after the capitals were in position. Indeed, if this were not the case, defenders of the pre-Conquest date of these towers would have to give up their position altogether. Unless we imagine that English artists used their memories to reproduce classical capitals which few of them are likely to have seen, we cannot imagine these capitals coming into existence otherwise than under Norman influence, and Norman influence after the Conquest. The most successful imitations of early Norman capitals are at Harpswell, where, as we have observed, the measurements of the tower are of a Norman and not of a Saxon type. At Caistor, unfortunately, the original belfry windows are gone, so that we cannot form any judgment as to the character of their capitals.