CHAPTER II

THE SHIELD AND ITS TREATMENT

Early Forms of Shields; Later Forms; Shields of Irregular Outline and Surface; The Filling of a Shield; Apparent _versus_ Absolute Uniformity; Modern Rules as to Proportion; The Use and Abuse of Quartering: its Origin and Growth; Differencing of Arms; The Scutcheon of Ulster; Diapering.

From these preliminary remarks we may pass to the practical consideration of the principles of heraldic art.

And first as to shields and their treatment.

The form of a shield is in itself entirely arbitrary and void of meaning. Although it varied from time to time, this was simply a matter of fashion, like the shape of an arch or the pattern of a window. Such changes must not, however, be overlooked, for it would be absurd in actual practice to use an ornate shield of the style of the fifteenth or sixteenth century for a lion of (say) the thirteenth century type, or to fill a shield of early form with charges characteristic of a later date.

During the twelfth century shields were more or less kite-shaped, like those that were actually used, but in the thirteenth century they began to be shorter and straighter across the top. Good examples of this type may be found on seals. In the aisles behind the quire of Westminster abbey church, the beautiful shields in the spandrels of the wall arcade, of a date not later than 1259, retain their rounded upper corners. (See figs. 2 and 15.) The next form, with the upper corners square (figs. 16, 17), came into vogue in the second half of the thirteenth century, and has continued always in use. Owing to the elastic way in which its curves can be slightly altered when required, it may safely be adopted in general practice. In the earliest examples the curves begin at the top, or just below, but later on they were so struck as to increase the area of the lower part of the shield in order to make more room for the charges. In some fourteenth century instances the sides continue straight nearly to the bottom, so that the shield is practically an oblong with rounded lower corners, like the shields of the royal arms on our coinage to-day (fig. 18 and pl. VI A). A tendency in the same direction is not uncommon throughout the fifteenth century. About the middle of the same century the fashion began to prevail, alongside the other, of representing a man's arms on the same irregularly shaped shield that he was wont to carry in the jousts. This is as wide at the bottom as the top, with its outline worked into curves, and has on the dexter, or right-hand side as borne, a deep notch for the lance to rest in during tilting; the top and bottom of the shield are often subdivided into three or more lobes or shallow curves. Good examples occur on seals and monuments, and some of the Garter stall-plates. (See pls. V A and B; VI B; XVII; and XXIII A.) Shields of a more ornate form are occasionally to be met with, like an example (fig. 19) on a brass at Stoke Poges of the date 1476, with graceful leaf-work curling over at the top and bottom. Shields similarly ornamented occur on the doorway of a citizen's house now built into the Guildhall at Norwich (fig. 20).

In the simpler forms the field of a shield in painted representations is invariably shown flat; but in carvings, and occasionally on seals, a slight convexity, or even concavity, is often met with, the artistic advantages of which it is unnecessary to enlarge upon. In some of the later ornate forms, like those described above, the incurved or engrailed edge is accompanied by a field worked with a series of ridges and furrows (figs. 21 and 23). The effect of this may be good, but there is a danger of carrying it to excess and so injuring the appearance of the charges. If the shield be well covered by the bearings on it, it is generally better to use one of simple form than one with an irregular outline and ridged surface; but there is, of course, no reason why both forms should not be used concurrently in architectural or other works, as they sometimes were of old.

The same principle as the ridging of a shield to relieve the plain surface was also applied to the ordinaries upon it. An early example may be seen upon the tomb of queen Eleanor at Westminster, which has the bends in the shields of Ponthieu ridged along the middle line. The shield borne by Brian FitzAlan (_ob._ 1302) in his effigy at Bedale has the alternate bars of his arms (_barry of eight pieces gold and gules_) treated in the same way. Another instance may be seen on the effigy of Sir Richard Whatton (_c._ 1325) at Whatton, Notts, in which a bend, though charged, is ridged. The shields on the tomb of Guy lord Bryen (_ob._ 1390) at Tewkesbury (fig. 22) furnish typical later examples, while during the fifteenth and early sixteenth centuries instances are as common as the curved and ridged shields described above, especially as regards crosses and saltires, as at St. Albans, the George Inn at Glastonbury (fig. 23), and elsewhere.

In monumental effigies the shield borne by a knight often has a convex or rounded surface (fig. 24), and in late fifteenth century and Tudor architecture otherwise flat shields sometimes have the middle swelled out, as on dean Gunthorpe's oriel at Wells, in a manner very popular in Renaissance work. (See figs. 111 and 195.)

A reference to a number of good ancient examples of heraldic shields or banners will disclose the care that has been taken to occupy the field, as far as possible, with whatever is placed upon it (figs. 25, 26, 27). A lion or an eagle, for instance, will have the limbs and extremities so spread out as to fill every available space; and the same will be found in every group or combination of objects capable of arrangement or extension.

Even with most unpromising combinations, or a group that cannot be extended or modified at all, or with a single charge like a fleur-de-lis, or ordinary such as a bend, (fig. 30), pale, or cheveron (pl. VIII A), a judicious adjustment of proportions, or some equally common-sense method, enabled a medieval artist to make his shield look well.

Another point that may be noticed in all old work is that in shields containing several similar objects no two are exactly alike. If the charges be, for example, three roses or three roundels or three lions (fig. 32), two will be placed in the upper and the third in the lower part of the shield. But the latter will often be somewhat larger than the others, and these, in turn, will differ slightly the one from the other as they do in nature. So, too, in a case like the three leopards of the King of England, whether displayed on shield or in banner, no two are exactly alike, but each differs somewhat from another in pose or in size (fig. 32). Even when the same charge is repeated many times, like the fleurs-de-lis in the old arms of France, any possible chance of mechanical monotony is avoided by a trifling variation in the shape of each, as in the shield of the King of France in the early series at Westminster (fig. 34).

Another fact is that in the old work lines and curves are hardly ever quite true, but drawn by hand instead of with pen or compasses. The modern artist, on the contrary, usually draws his lines and curves with mechanical precision; his charges are exact copies one of another; the fact that they do not fill the field (_pace_ the royal arms on the coinage) is to him quite unimportant, and the final result is that under no circumstances will his work look well. Even in old stencilling a pleasing effect never seen in modern work of the kind was produced through a not too rigid adherence to a regularity of application.

Another cause of the bad effect of much modern heraldry is the unnecessary adherence to the rules laid down in some of the textbooks and manuals as to the relative widths of ordinaries and subordinaries. The old heralds certainly did not fetter themselves with such shackles. A cheveron, a bend, a fesse, or a cross was drawn of the best proportion to look well (figs. 35, 36). If charged it would be wider than when plain. If placed between charges it was drawn narrower, if itself uncharged, and thus took its proper relative position with regard to the size and arrangement or the charges. So, too, with a border; if uncharged or merely gobony (_i.e._ formed of short lengths of alternate colours) or engrailed, it was drawn very narrow, and even if charged it was not allowed much greater width (figs. 38, 39). It thus never unduly encroached upon the field or other contents of the shield, and yet remained an artistic addition in itself. The curious bordering known as the tressure, which is almost peculiar to Scotland, and familiar to us through its occurrence in the shield of our Sovereign, is drawn sufficiently narrow in all good examples to leave ample room for the ramping lion it fences in, and its frieze of fleurs-de-lis is formed of a good number of flowers, instead of the eight considered sufficient in the royal arms of to-day. Even a chief, if necessary, was enlarged from the 'less than one-third of the shield' of to-day to the one-half of it, or even more, as may be seen in some of the examples of the arms of the monastery in the abbey church of Westminster, or in those of the town of Southampton.

Another feature of early heraldry which it is well to bear in mind is the sparing use of what is known as quartering, or the method of combining in one shield the arms of two or more persons or families. One of our oldest instances of this occurs on the tomb of Queen Eleanor, the first wife of King Edward I, at Westminster, and shows her paternal arms of Castile and Leon so arranged (fig. 40). Another early example occurs in the Great Roll, _temp._ Edward II, where the arms of Sir Simon Montagu (_ob. c. 1316_), _silver a fesse indented gules of three indentures_, are quartered with _azure a gold griffin_. So long as the shield contained only four quarters, with the first and fourth, and the second and third, respectively alike, the effect was often good, as in the cases just noted, or in the beautiful arms of France and England combined used after 1340 by King Edward III (fig. 41). There are also many examples, as in the well-known bearings of the Veres and of the Despensers, where a quarterly disposition of the shield forms the basis of the arms. But when, as became common in the fifteenth century, quarters were multiplied or subdivided, the artistic effect of the old simple shield was lost or destroyed. As the principle was further extended, especially in Tudor and Stewart times, the result became more and more confused in appearance, until the field resembled rather a piece of coloured patchwork than a combination of various arms all more or less beautiful in themselves.

The origin and growth of these combinations, which actually are perfectly lawful and proper, and yet often quite accidental, can easily be illustrated by a few typical examples.

In 1382 King Richard II, who used the same arms as his grandfather, a quarterly shield of Old France and England, married Anne of Bohemia, daughter of the Emperor Charles IV. As her shield was also a quartered one, the combined arms of the king and his queen, as shown upon her seal, formed a shield of eight quarters (pl. VII A). This was further complicated through the later assumption by King Richard of the arms assigned to St. Edward (fig. 43), a cross between five birds; and the eight-quartered shield with this clumsy addition at one side may be seen on the Felbrigge brass.

These arms of St. Edward were used for a time duly 'differenced' in conjunction with his own quarterly arms by Henry of Lancaster, afterwards King Henry IV, and are impaled with those of his wife, Mary de Bohun, on his seal (1399) as duke of Hereford. Artistically the lop-sided effect so produced is quite unhappy.

Many fifteenth century shields show forth, by the simple quartering of a man's arms with those of his wife or his mother, his succession or summons as a lord of parliament, or his inheritance of great estates.

But this simplicity was gradually destroyed when the added quartering was itself quartered, as in the arms of Richard Nevill earl of Salisbury (see pls. XVII A and XXII B), or the quarterings were all different, as in the case of Humphrey Stafford duke of Buckingham. When but a year old he succeeded his father as earl of Stafford, and on his mother's death he became earl of Buckingham, Hereford, Northampton, Essex, and Perche! These dignities are duly displayed in the quarterings of his arms on his seal, as follows: 1. The quartered arms of his mother, for the earldom of Buckingham. 2. Bohun of Hereford. 3. Bohun of Northampton. 4. Stafford (fig. 44).

When Henry duke of Buckingham succeeded in 1460 to all the dignities of duke Humphrey his grandfather, he wisely elected, by the advice of the kings-of-arms, to drop the above quarterings, and to use only the arms of his great-grandmother, who as sister and heir of Humphrey duke of Gloucester and earl of Buckingham bore _France and England quarterly within a border silver_.

About 1433 Margaret, daughter of Richard Beauchamp earl of Warwick, was married to John Talbot earl of Shrewsbury, and she thereupon had a beautiful seal engraved, with two large shields or arms hung side by side by their straps from a ragged staff, the badge of her father's house (pl. XXVII B). This charming composition is, however, quite spoilt through the complicated treatment of the shields. One of these bears the arms of husband and wife conjoined, the other those of the lady's father. The earl of Warwick's shield is a quartered one of Beauchamp and Newburgh, with a small superimposed scutcheon. The earl of Shrewsbury's arms also consisted of four quarters, to which his wife added her four (omitting the scutcheon), and thus made a patchwork of eight.

A more remarkable and equally accidental case may be illustrated by the brass of Sir Humphrey Bourchier (1471) in the abbey church of Westminster.

This displays four shields: one has the arms of Bourchier quartering Lovain and impaling the quarterly arms of Berners; and another, the six quarterings of Sir Humphrey's wife, Elizabeth Tylney. In a third shield these are quite properly impaled, with a resultant of fourteen quarters. In the fourth shield these are quartered together, and so produce a dreadful confusion of twenty-eight quarters! It is not necessary here to show how these shields might have been simplified in themselves, but from the artistic standpoint there cannot be any doubt that the two first should at least have been kept separate. The many other examples to be found in the illustrations of this book will serve as useful reminders of the greater advantage artistically of simpler treatment.

It is moreover well to remember that in the majority of cases there is not the least need in actual work to produce a great many quarterings in a shield. In numerous examples, especially in the sixteenth century and later, they were assumed merely for display, and to reduce them to a reasonable few is often a most desirable thing.

It is difficult without knowledge of individual cases to lay down any definite rules for dealing with quarterings, but there can be no question that in general a shield looks best without any at all. In the case of a man with a compound name or title, who represents more than one family or dignity, it would be legitimate to add a quartering on that account, but only of the actual arms of the family or dignity represented. It is however so hard to draw a line or to restrain the wishes of clients that the fifteenth century example of Henry duke of Buckingham should ever be borne in mind.

As soon as the principle of hereditary descent of armorial bearings became established, the necessity arose of making some slight difference between the arms of a father and those borne by his sons. This was usually done by adding to the paternal arms such more or less unobtrusive device as a label, or narrow border, or a small charge like a crescent or a molet.

The lord John of Eltham, son of King Edward II, bears upon his tomb at Westminster a beautifully carved shield of the arms of England differenced by a border of France; and one of the sons of King Edward III, Thomas of Woodstock, differenced his father's arms by a silver border, as at an earlier period did Edmund earl of Kent, the youngest son of King Edward I.

The label is a narrow band with long pendent strips or pieces, usually three, but sometimes four or five in number, placed upon and across the upper part of a shield (fig. 45). It is now used to distinguish the arms of an eldest son from those of his father, but this was not always the rule, and younger sons of King Henry III and King Edward I, and at least three of the sons of King Edward III, besides the Prince of Wales, bore distinctive labels for difference. Anciently, the label was very narrow, and the pendent pieces of equal or nearly equal width throughout, even when charged with devices, as they sometimes were. The colour was also a matter of choice. The first three Edwards, during their fathers' lifetime, successively bore blue labels, sometimes of three, sometimes of five pieces, while the younger brother of King Edward I, Edmund earl of Lancaster, used a label of France (blue with gold fleurs-de-lis) of four pieces, and Thomas of Brotherton, second son of King Edward I, a silver label of three pieces.

In the case of the sons of King Edward III, the Prince of Wales bore at first a silver label of five and later of three pieces; Lionel duke of Clarence seems to have borne at one time a gold label with a red cross on each piece for Ulster, and at another a silver label charged on each piece with a red quarter for Clare; John of Gaunt duke of Lancaster bore an ermine label for his earldom of Richmond (pl. II); and Edmund duke of York a silver label with three red roundels on each piece (pl. XXI B). The rolls of arms furnish instances of labels of all colours, and with pieces charged with various devices, such as leopards, eagles, castles, martlets, etc.

Differencing with labels was likewise extended to crests, and a good example may be seen on the monument of Edward prince of Wales (_ob._ 1376) at Canterbury (fig. 46), as well as in fig. 139.

In modern heraldry the label is often drawn unduly wide, with short and ugly wedge-shaped pieces hanging from or sticking on to it, and sometimes it does not even extend to the sides of the shield. The result is that instead of its being a comparatively unobtrusive addition to the arms the label becomes unduly conspicuous and void of all artistic effect.

The old way of differencing by the addition of a crescent, molet, or similar device was generally carried out in quite an artistic fashion on account of the care taken to place the device agreeably, a favourite position being on the principal ordinary or charge of the arms.

Many cadets of the great family of Nevill, for example, differenced the arms of their house, _gules a saltire silver_, by placing the device on the middle of the saltire, and some of the Beauchamps placed the differencing mark on the fesse of their arms. In other cases the device was placed in the upper part of the shield, or in some other such point where it would least interfere with or be confounded with the charges.

One of the most difficult differences an artist has to contend with to-day is the silver scutcheon with a red hand which is placed upon the arms of baronets. Its position of course varies, and may often be altered with advantage, and it looks all the better if drawn not unduly large and with a simple heater-shaped shield. But some artists wisely leave it out altogether.

In the case of all devices introduced as differences it will generally be found advisable to draw them to a somewhat smaller scale than the charges already in the arms.

In many ancient heraldic shields, especially in painted glass, and to a lesser extent in carved work and on seals, the plain uncharged surfaces of the field or ordinaries are relieved by covering them with the purely ornamental decoration called diapering (figs. 45, 48). An early instance in relief occurs on the shield of the effigy in the Templars' church in London usually ascribed to Geoffrey de Magnavilla; and another delicately sculptured example of later date is to be seen on the Vere effigy in Hatfield Broadoak church in Essex (fig. 47). Several fine instances of painted diapering will be found in Stothard's _Monumental Effigies_. This beautiful treatment has, happily, been largely revived of late years by the glass painters, who use it quite successfully, probably from the ease with which in their case it can be applied. Modern carvers use it very sparingly, and this perhaps is as it should be, for diapering needs to be done with great skill in sculpture to look well. A careful study therefore of old examples is advisable, in order thoroughly to understand the principles of its application.

Some of the finest diapered shields in carved work occur in the spandrels of the splendid monument of the lady Eleanor Percy in Beverley minster (figs. 49, 50). Good instances are to be found on seals, and a number of these are here illustrated in order to show the proper treatment of diapering. (See pls. VIII, XII, and XXVII A.)

It is of course to be borne in mind that diapering is merely a surface decoration, and it must not on any account be emphasized by any difference of colour from that of the field or ordinary it relieves, nor must it be treated with such prominence as to render it liable to be mistaken for a charge or charges.

Diapering can be represented effectively in embroidered work by the use of flowered or patterned damasks, as may be seen in the banners in St. Paul's cathedral church in the chapel of the Order of St. Michael and St. George.