Among the studies for the head of the Virgin, I may mention as most important a drawing on green paper in silver-point, in the Duke of Devonshire's collection at Chatsworth (see our pl. vi.). In this she is represented side by side with the little S. John, looking from left to right, in an exactly opposite direction to that of the picture. The type, over-slender and affected, is far from attractive, and differs altogether from that finally adopted. But that this head was a study for the Virgin of the Rocks is proved by the presence of the

(Royal Library, Turin.)

little S. John, reproduced almost exactly from the drawing in the Louvre.

Having thus established the relation of

the Chatsworth drawing to the Virgin of the Rocks, we are further enabled to connect a head of the Virgin in the Christ Church collection at Oxford (p. 171), with the picture. In type and technique this drawing

is almost identical with

that at Chatsworth.1

I may add that, differing altogether from Herr Müller-Walde (Fig. 8), I consider the head of a young woman on green paper, in the Uffizi, closely akin to the head of the Virgin in

1 This connection has escaped Herr Müller-Walde, who assigns the date 1472-1473 to the Christ Church study. (Fig. 9, pl. xliii.) We must, in view of the demonstration in the text above, antedate it by some six or eight years. As to the laborious theory built up by Signor Morelli on the Christ Church drawing, which he ascribes to his favourite, Bernardino dei Conti, it is overthrown at once by the mere fact that this drawing was a study for the Virgin of the Rocks, and that in execution it shows an absolute identity with other drawings by Leonardo. It is not improbable that the head of a woman in the Borghese Gallery (Müller-Walde, Fig. 7) may also have been a study for the picture, if indeed this drawing is really by Leonardo.

the picture. It has the same short but firmly modelled nose, the same straight lips, the same somewhat square chin.1

We may now briefly mention the studies for the Infant Jesus.

The Louvre owns three, in silver-point heightened with Chinese white, on that greenish paper Leonardo seems to have specially affected during his first Florentine period. They are all of the Child's head, and show it in profile; he looks before him, while, in the picture, he turns to look at his mother. Note, however, that whereas in the first the face is in sharp profile, in the other two the artist tries the effect of a profil perdu" (i.e., less than a full profile). Dr. Richter (vol. i., p. 345) questions the authenticity of the principal drawing (no 383 in M. Reiset's catalogue), which he holds to be a copy of later date. But I am unable to share his views on this point. Herr Müller-Walde, on the other hand, describes the drawing as "herrlich" (superb).

A smaller, but more complete study of the same head, with the shoulders and part of the breast added, is in the Royal Library at Windsor (Richter, pl. xliv.). It is a very realistic drawing, the expression of the face curiously old and prescient. It is noticeable that it is in red chalk, a medium never used by Leonardo's predecessors, and infrequently by himself till a comparatively late period of his career. Nothing short of Richter's authority, therefore, would induce me to accept the authenticity of this study, the earliest in date of Leonardo's drawings in red chalk.

Another study for the Child, seated, and leaning on one hand, an angel's head beside him, was published by Gerli (pl. xix.).

Finally, a pencil drawing of a child's head, touched with Chinese white, in the Chatsworth collection, is also supposed to be a study for the picture.2

We may now pass on to the studies for the little S. John. A sketch for the head, three quarters to the front, is to be found in the Vallardi collection, in the Louvre (Braun, no. 170). It is drawn in silver-point, on greenish paper: (Richter, vol. i., 342). This head

1 I only know the grisaille sketch for the head of the Virgin in the Holford collection

by Rio's mention of it in L'Art Chrétien (vol. iii., p. 81).

2 Waagen, Treasures of Art in Great Britain, vol. iii., p. 353.

served Raphael as his type for a whole series of Infant Saviours. The same head re-appears in a drawing in the Duke of Devonshire's collection, also on green paper, side by side with a head of the Virgin. (See our pl. vi.)

Two drawings in the Mancel gallery, in the Hôtel de Ville at Caen, to which my attention was drawn by M. Léopold Mabilleau, and for photographs of which I am indebted to the learned keeper of the gallery, M. Decauville-Lachenée, are studies for the little S. John and the Infant Jesus. A long interval, however, perhaps several years, seems to have divided these studies from the finished work. As his habit was, the artist, before sitting down to his easel, submitted his various figures to a laborious process of adaptation. Thus, he made the profile head considerably younger in the picture; from a boy, the child became an infant. He also reduced the masses of hair to normal proportions, and softened the expression of the little S. John.

German critics, from Passavant and Waagen to Herr Müller-Walde, have contested the authenticity of the Virgin of the Rocks from time to time.1

Setting all patriotic considerations aside, I cannot but maintain the Louvre picture to be one of those in which the master's genius manifests itself most gloriously. Allowances must, of course, be made for the unhappily numerous repaints, and the blackening of the shadows, a defect aggravated by the thick yellow varnish that overlies the surface. Granted that the composition has not achieved the breadth and grandeur of the Last Supper, or the suavity of the S. Anne, yet it shows us Leonardo as his own precursor.

A replica of the Virgin of the Rocks was bought in 1880, at the considerable price of £9,000, for the English National Gallery, which claims in this example to have acquired the true original by Leonardo. The replica, which came from the Suffolk collection, was bought in

1 Quite recently, M. Strzygowski pronounced the Virgin of the Rocks, in the Louvre, a bad copy! (Jahrbuch der kgl. Kunstsammlungen, 1895, p. 165.) See also Sir E. Poynter's article in the Art Journal, 1894, p. 229-232. But cf. Signor Frizzoni (Gazette des Beaux Arts, 1884, vol. i., p. 235), Herr Koopmann (Repertorium für Kunstwissenschaft, 1891, p. 353-360), Dr. Richter (Art Journal, 1894, pp. 166-170, 300-301), and various other foreign critics, who all uphold the Louvre picture.