Leonardo da Vinci Read online

  The general effect of the Uffizi Annunciation is obscured by its condition. The whole surface is extremely dirty and the draperies are discoloured by old retouches and congealed varnish. But even allowing for this, the execution must have been rather uneven and in places relatively coarse; and the red of the angel’s robe can never have borne much relation to the pink of the Virgin’s skirt. Incidentally, we must notice that the angel’s wing has been lengthened to canonical proportions by a very crude overpainting and hangs like a brown smear above the enchanting landscape to the left. The original short wings are directly painted from the wings of a bird, and fit the angel’s shoulders with convincing naturalism. The angel’s head (Pl. 3) is nearer to the traditional quattrocento type than anything else in Leonardo’s work, and at first sight has some of the tameness of a Ghirlandajo. Steadier contemplation will disclose a subtlety of outline and modelling, especially round the nose and eye, and a greater technical confidence than in the angel in the Baptism. The Madonna’s head (Pl. 4), on the other hand, is very like the earlier angel. The features are not felt as part of the structure of the face, but are drawn on it, and if we compare it with the Mona Lisa we may begin to realize the immense labour which Leonardo devoted to studying the science of his art.

  In spite of certain faults, the Annunciation remains a lovely and original picture, in which shortcomings of composition are outweighed by beauties of detail and of mood. No other work of Leonardo does so much to support Vasari’s account of his early sympathy with nature. Other painters of his century had painted nature decoratively or accurately, but none, unless it were the blessed Angelico, had interpreted her moods, and used them to set the emotional key of a picture. In the Annunciation the black trees silhouetted against the grey evening sky are one of those effects which first, in our childhood, made us feel the poetry and, as it were, the closeness of nature. It is an effect which, from its easy appeal, has been often abused, but Leonardo has deprived it of all sentimentality. The distant view of a river with high hills beyond is more in the Flemish fashion than any other of Leonardo’s backgrounds, but the lesser landscape to the left of the angel, dimly perceptible through the dirt and varnish, has the personal vision of the Uffizi drawing. Even more similar to his mature work are the flowers in the foreground. The artists of the quattrocento had spread their flowers like verdure tapestry, or drawn them in isolation, like botanical specimens, but Leonardo has given to his flowers and grasses something of the turbulence which he felt to be the essence of nature. They twist and surge like little waves over the space between the angel and the Madonna, giving vitality to what would otherwise have been a dead area in the composition.

  Soon after the Annunciation Leonardo must have painted the portrait of a lady in the Liechtenstein Collection (Pl. 5). It is based on the same sequence of tones as the angel’s head—the pale flesh, the dark mass of foliage, the luminous grey sky; but the head is more firmly modelled than the Virgin Annunciate, and must be slightly later. The Liechtenstein portrait is of an exquisite melancholy beauty, far outside the range of Verrocchio and beyond the power of Credi; and personally I have little doubt that it is the portrait of Ginevri Benci, mentioned in Vasari and the Anonimo Gaddiano,{8} although Vasari says that this was painted during Leonardo’s second Florentine period. It seems to represent a lady called Ginevra, for not only does a bush of juniper{9} form the background of her head, but on the reverse of the panel is a sprig of juniper encircled by a wreath of laurel and palm. Most Florentine portraits of women were painted to celebrate their marriages. Ginevra Benci was married in January 1474, and this is a possible date for the picture on grounds of style.

  We can guess by its unusual shape that the picture has been cut at the bottom, and this fact is confirmed by the truncated wreath at the back. By completing the curve of the wreath and allowing for decorative ribbons, etc., we can calculate that the amount cut off must have been at least nine centimetres, which would give it the classical proportion of 3 to 4. This would also make it large enough to contain the lady’s hands, and Bode, who was the first scholar to give the picture full attention, suggested that these hands were known to us in a famous and beautiful silverpoint drawing at Windsor (Pl. 6). The hands at Windsor, with their long-jointed fingers, are like enough to those of the Virgin in the Uffizi Annunciation, and the style of the drawing is just possible for 1474, although firmer and more masterly than anything else of that date which has survived. Assuming that the drawing was for the Ginevra, we see that the fingers of her right hand would have been touching the lower laces of her bodice; and in the picture these laces are in fact repainted. Unfortunately, we can tell from the back that the bottom of the picture has been damaged and a new piece has been added, so that an X-ray would not reveal the vanished fingers.

  In spite of mutilation the Ginevra Benci is the best preserved of all Leonardo’s early pictures, and shows most clearly his intentions at this period. Areas of light and dark are strongly contrasted, but within the light oval of the face there is very little shadow, and the modelling is suggested by delicate gradations of tone, especially in the reflected lights. We see a similar treatment of form in Desiderio’s low reliefs, controlled by the same sensibility to minute variations of surface. There are passages, such as the modelling of the eyelids, which Leonardo never surpassed in delicacy, and here for once he seems to have had none of that distaste for the medium which we can deduce from his later paintings, no less than from contemporary descriptions of his practice. Most ingenious is the way in which all the light areas except the face are given broken irregular contours by the juniper leaves, or the shimmering water, so that the outline of the lady’s cheek and brow dominate the design. There is a similar contrast in texture between the beautiful curves of her ringlets and the stiff, spiky character of the juniper. But all these technical devices are subordinate to the feeling of individual character with which Leonardo has been able to charge his portrait, so that this pale young woman has become one of the memorable personalities of the Renaissance.

  Assuming that the Ginevra was painted in 1474, how did Leonardo spend the next four years? There are no documents bearing on his work till 1478, no drawings and only one picture which seems to belong to the period. At twenty-five years of age, he cannot have been obscurely devilling for Verrocchio, although we know that he lived with Verrocchio till 1476; and he had not yet begun the scientific studies which, in his later life, account for intervals in the sequence of his paintings. Presumably Leonardo, like other young men with great gifts, spent a large part of his youth in what is known as doing nothing—dressing up, talking, taming horses, learning the lute, learning the flute, enjoying the hors d’œuvres of life, till his genius should find its true direction.

  There is one picture which can be dated in this period with some show of evidence, the Virgin with the vase of flowers in the Munich Gallery. It is connected in many ways with the studio of Verrocchio. Credi did a drawing and picture of the same model in the same costume, and an almost identical pose was used in a composite production of Verrocchio’s shop, the altar-piece in Pistoja Cathedral. Perhaps the fact that the Munich Madonna was little more than a workshop commission accounts for the absence of most of those qualities which we value in Leonardo’s other work. The picture is hard to judge in its present condition. The Virgin’s head has been entirely repainted in a medium containing too large a quantity of oil, resulting in a craquelure so unlike that of any Italian technique that Morelli believed the whole picture to be a Flemish copy. The same medium was used to rework the shadows, especially round the baby’s head, of which a different outline is faintly perceptible. There are many other damages, and it may well be asked on what grounds the picture can be ascribed to Leonardo. The answer is that all the surviving parts are wholly characteristic. The Virgin’s plaited hair, and her left hand, large parts of her drapery, and the flowers in the vase at her side, are all painted in exactly the same style of the Uffizi Annunciation, and they combine to give the pict
ure as a whole a quality of form and colour which is unlike anything else of the period. For further confirmation there is the panorama of mountains seen through the arched windows of the background, which is closer to his mature type of landscape than the background of the Annunciation. Already Leonardo has felt it necessary to turn the horizontal line of the background into a series of verticals by rows of precipitous mountains. Like the flowers in the Annunciation these Gothic pinnacles are a negation of repose, a refusal to allow that anything in nature should be devoid of movement. It is instructive to compare this charmless picture with the exquisite little Virgin and Child from the Dreyfus Collection, now in the National Gallery, Washington, which must be a very early work by Lorenzo di Credi. Both were painted from the same studio model and perhaps at the same time. The Dreyfus Madonna is executed with a delicacy and minuteness of touch which resists enlargement up to ten times its original size. It is the work of a born prize-winner. Where the Munich Madonna has the unpleasant vitality of immature genius, the Dreyfus Madonna has the amiable complacency of the craftsman.

  In January 1478 we have the first record of Leonardo being given an important independent commission, the altar-piece in the Chapel of St. Bernard in the Signoria, Florence. A first payment was made, but for some unknown reason the work was never delivered, and in 1483 the commission was given to Ghirlandajo. The Anonimo Gaddiano says that it was finally executed by Filippino Lippi, who worked ‘on the design of Leonardo’, but Filippino’s altar-piece in the Uffizi, dated 1485, shows no connexion with Leonardo either in spirit or design. To the same year belongs a drawing of two heads with a fragmentary inscription which mentions the town of Pistoja and states that in the autumn of 1478 Leonardo had begun two pictures of the Madonna—...bre 1478 inchominciai le 2 Vergine Marie. One of these must almost certainly be the so-called Benois Madonna, now in the Hermitage (Pl. 8). The Benois Madonna, the latest of the four pictures painted in the 1470s, is the one which is most generally accepted by the critics. Indeed, it is the unquestionable authenticity of this picture which has driven all but the most idealistic critics to accept the other three. The Benois Madonna does not owe this position so much to its intrinsic merits as to the fact that there exist several drawings for the composition which are undeniably by Leonardo. Here, for the first time, we are able to study his art with a method which is always illuminating, namely by comparing his drawings with his finished pictures. And immediately we find evidence of a conflict between spontaneity and perfection. With the drawings for the Benois Madonna we may also consider a rather more numerous series of drawings for a Virgin and Child, now lost, in which the Child is playing with a cat or tame weasel (Pl. 10), and by treating these Madonna studies as a single group we can take a wider basis for our generalizations. These studies are so fresh and natural that they are now amongst the most popular of all Leonardo’s drawings. They show, as nothing else in his work, a direct and happy approach to life; and they show his matchless quickness of vision, which allowed him to convey every movement or gesture with the certainty and unconscious grace of a great dancer performing a familiar step. In one sense the word ‘certainty’ does not apply to these early drawings. Technically they are still experimental. The line is recklessly free; sometimes they are scrawled over and covered with blots and washes of sepia so that they resemble drawings of the seventeenth century—Rembrandt or Guercino. But the general effect is one of a graceful speed and of a hunter’s certainty of eye. The largest and most beautiful of this series is a study in the Louvre of the Virgin holding out a plate of fruit to the Infant Christ, in a pose close to that of the Benois Madonna (Pl. 9). It is drawn with a few rapid and summary strokes, as though Leonardo’s whole aim had been to note down action; and yet he has achieved a perfect composition. The rhythmic relation of the two heads is as spontaneous and as inevitable as the relation between two bars of Mozart.

  In the Benois Madonna this unity of effect is lost. What has come between the sketch and the picture? To answer this question we must glance for a moment at the two traditions which divided Florentine art of the fifteenth century. One of these is the tradition of linear grace and fancy, the tradition of Lorenzo Monaco, Fra Filippo, and Botticelli; the other is the tradition of scientific naturalism founded by Masaccio and kept alive in Leonardo’s own day by his master Verrocchio. Inevitably these traditions overlapped. Botticelli had a phase of naturalism; Pollajuolo belongs as much to one side as the other. Leonardo, as we see from his drawings, belonged by nature to the first group. But by training he was of the second, and his powerful intellect led him to sympathize with the scientific approach. Thus between the sketch and the picture he was forced to attempt a complete change of mood and to adapt his fleeting visions to the severe standards of academic Florentine art. We have plenty of evidence how this adaptation was effected. First come the studies of action, just described, the motives which are later to be used in the picture. Then come what I may call the diagrams. These are usually quite small drawings, done from memory, and are syntheses of the most satisfactory motives. One of the diagrams for the Benois Madonna has survived and is typical in its severity; typical, also, in that Leonardo has followed it closely. So far I imagine that his powers were perfectly uninhibited, and had he been content to rely on suggestion rather than complete statement, like Rembrandt, or to accept the formulae of his time, like Raphael, he might have been a prolific painter. But he would neither improvise nor conform. He determined to work out every detail according to his own standard of perfection, a standard which included scientific accuracy, pictorial logic, and finish. To achieve this ideal, the period between the sketch or diagram and the finished picture was one of intense intellectual effort, in which every detail was studied and assimilated to a satisfactory form.

  This stage, too, we know from drawings, highly finished studies of detail, which often tell us more about Leonardo’s final intentions than the actual picture; because by the time he began to paint, constant labours and anxieties had so deprived him of all appetite for his subject that his pictures were either left unfinished, or, as with the Benois Madonna, were carried through without that vitality, that spontaneous rendering of action, which was the original motive of his whole conception.

  But it is a mistake to look in the Benois Madonna for the charm and freshness of the quattrocento. Rather, we must think of it as a change-ling from the high Renaissance, an immature sample of that intellectual, classical style which Leonardo was to evolve while painting the Last Supper. How near he came to this style as early as 1480 we can see by comparing the drapery of the Madonna’s right leg and thigh with the studies of drapery made for the Louvre Virgin and St. Anne about thirty years later (p. 62). The system of folds is almost exactly the same, and there can be no doubt that Leonardo, in the most mature and complex of all his works, used as a point of departure this design of his youth. The composition of the Benois Madonna, which has an air of perfect naturalness and simplicity, is remarkably original, since it is based on a scheme of diagonal recession unusual in painting before Leonardo. Even the absence of ornament, the contempt for mere decoration which is one of its least attractive qualities, is a presage of the austere and elevated style which was to become fashionable some forty years later.

  For all these reasons the Benois Madonna is a great advance on the Madonna with the vase of flowers at Munich; but it suffers from many of the same defects. The condition is equally bad in a less obvious way. Under the discoloured varnish it is covered with small stippled touches which deprive it of all transparency of handling. The Virgin’s teeth, for example, are so entirely obscured by dirt that one distinguished critic, judging from a photograph, wrote of her as toothless. As the eye gradually penetrates these obstructive layers, her whole head takes on some of the vitality of the original drawings, and it is possible that the deadness which I have just referred to is partly due to condition. But parts of the composition can never have been happy: the baby was always monstrous and the drapery of the sleeve lab
oured. Finally, the one redeeming feature of the Munich picture, the landscape seen through a window, is absent from the background of the Benois Madonna. The window itself, lacking a central transom, is ugly enough, and without a landscape it is really painful. Perhaps a landscape once existed and has been overpainted, with the result that a large patch of light sky puts the Virgin’s face out of tone and destroys the unity of focus—the very mistake which Leonardo so skilfully avoided in the portrait of Ginevra Benci.

  It is possible to reconstruct at least two more pictures of the Virgin and Child dating from this period. One of these, the Virgin almost in profile, known from the Madonna Litta in the Hermitage, which is either a copy or a ruined original, is discussed in my next chapter. The other is known to us from a drawing on the famous sheet at Windsor which also contains some of the earliest of those characteristic profiles, the pretty boy and the toothless Roman warrior, which were always to be the first scribbles to flow from Leonardo’s pen. The Virgin is represented half kneeling on her left knee, holding the Child on her right, with the infant St. John standing close beside her. The drawing is summary, but we know that it represents something like Leonardo’s final version of the subject, as the group of the Virgin and Child is reproduced almost exactly in a picture by Andrea da Salerno at Naples. This proves that Leonardo carried the idea much further than a sketch, perhaps to a painting, certainly to a cartoon. And what an immensely influential composition it was! Mr. Berenson has shown that it is probably the earliest representation of the Virgin and Child to include the infant St. John, and even if earlier instances could be found, it is certain that the use Leonardo has made of this iconographic motive is in the highest degree original. He has discovered the secret of that pyramidal composition which became an academic dogma of the high Renaissance; for the infant St. John, standing beside the seated Virgin, gives just that weight and balance to the base of the pyramid which the Virgin and Child alone would otherwise have lacked. Having invented this motive and used it once, Leonardo abandoned it, with all its permutations and combinations, to be worked out by Raphael. The Belle Jardinière, Madonna with the Goldfinch, Madonna of the Meadows, and the Esterhazy Madonna are variations on a theme by Leonardo: which, as any musician knows, does not make them less beautiful or personal.