slope to go up, the first ascent of the purgatorial mountain. The similes by which he illustrates the steepness of that ascent are all taken from the Riviera of Genoa, now traversed by a good carriage road under the name of the Corniche; but as this road did not exist in Dante's time, and the steep precipices and promontories were then probably traversed by footpaths, which, as they necessarily passed in many places over crumbling and slippery limestone, were doubtless not a little dangerous, and as in the manner they commanded the bays of a sea below, and lay exposed to the full blaze of the south-eastern sun, they corresponded precisely to the situation of the path by which he ascends above the purgatorial sea, the image could not possibly have been taken from a better source for the fully conveying his idea to the reader: nor, by the way, is there reason to discredit, in this place, his powers of climbing; for, with his usual accuracy, he has taken the angle of the path for us, saying it was considerably more than forty-five. Now a continuous mountain slope of forty-five degrees is already quite unsafe either for ascent or descent, except by zigzag paths; and a greater slope than this could not be climbed, straightforward, but by help of crevices or jags in the rock, and great physical exertion besides.

§ 14. Throughout these passages, however, Dante's thoughts are clearly fixed altogether on the question of mere accessibility or inaccessibility. He does not show the smallest interest in the rocks, except as things to be conquered; and his description of their appearance is utterly meagre, involving no other epithets than "erto" (steep or upright), Inf. xix. 131, Purg. iii. 48, &c.; "sconcio" (monstrous), Inf. xix. 131; "stagliata" (cut), Inf. xvii. 134; "maligno" (malignant), Inf. vii. 108; "duro" (hard), xx. 25; with "large" and "broken" (rotto) in various places. No idea of roundness, mas

66

siveness, or plesant form of any kind appears for a moment to enter his mind; and the different names which are given to the rocks in various places seem merely to refer to variations in size: thus a "rocco " is a part of a "scoglio," Inf. xx. 25 and xxvi. 27; a "scheggio" (xxi. 69 and xxvi. 17) is a less fragment yet; a "petrone," or sasso," is a large stone or boulder (Purg. iv. 101, 104), and "pietra," a less stone,-both of these last terms, especially "sasso," being used for any large mountainous mass, as in Purg. xxi. 106; and the vagueness of the word “monte" itself, like that of the French "montagne," applicable either to a hill on a post-road requiring the drag to be put on, or to the Mont Blanc, marks a peculiar carelessness in both nations, at the time of the formation of their languages, as to the sublimity of the higher hills; so that the effect produced on an English ear by the word "mountain," signifying always a mass of a certain large size, cannot be conveyed either in French or Italian.

§ 15. In all these modes of regarding rocks we find (rocks being in themselves, as we shall see presently, by no means monstrous or frightful things) exactly that inaccuracy in the medieval mind which we had been led to expect, in its bearings on things contrary to the spirit of that symmetrical and perfect humanity which had formed its ideal; and it is very curious to observe how closely in the terms he uses, and the feelings they indicate, Dante here agrees with Homer. For the word stagliata (cut) corresponds very nearly to a favorite. term of Homer's respecting rocks "sculptured," used by him also of ships' sides; and the frescoes and illuminations of the Middle Ages enable us to ascertain exactly what this idea of cut" rock was.

§ 16. In Plate 10 I have assembled some examples, which will give the reader a sufficient knowledge of mediæval rock-drawing, by men whose names are known.

They are chiefly taken from engravings, with which the reader has it in his power to compare them,* and if, therefore, any injustice is done to the original paintings the fault is not mine; but the general impression conveyed is quite accurate, and it would not have been worth while, where work is so deficient in first conception, to lose time in insuring accuracy of facsimile. Some of the crags may be taller here, or broader there, than in the original paintings; but the character of the work is perfectly preserved, and that is all with which we are at present concerned.

Figs. 1 and 5 are by Ghirlandajo; 2 by Filippo Pesellino; 4 by Leonardo da Vinci; and 6 by Andrea del Castagno. All these are indeed workmen of a much later period than Dante, but the system of rock-drawing remains entirely unchanged from Giotto's time to Ghirlandajo's;-is then altered only by an introduction of stratification indicative of a little closer observance of nature, and so remains until Titian's time. Fig. 1 is exactly representative of one of Giotto's rocks, though actually by Ghirlandajo; and Fig. 2 is rather less skilful than Giotto's ordinary work. Both these figures indicate precisely what Homer and Dante meant by "cut" rocks. They had observed the concave smoothness of certain rock fractures as eminently distinctive of rock from earth, and use the term "cut" or "sculptured" to distinguish the smooth surface from the knotty or sandy one, having observed nothing more respecting its real contours than is represented in Figs. 1 and 2, which look as if they had been hewn out with an adze. Lorenzo Ghiberti preserves the same type, even in his finest work.

Fig. 3, from an interesting sixteenth century MS. in the British Museum (Cotton, Augustus, a. 5), is characteristic of the best later illuminators' work; and Fig. 5,

* The references are in Appendix I.