The pupil and biographer of Piette, l'Abbé Henri Breuil, observes that his scheme of art evolution is exact along its main lines.(20) It is true that human sculpture appears for the first time in the lower Aurignacian, that it survives the Solutrean, and even extends into middle Magdalenian times, but this enormous period cannot be placed in one archæological division as Piette supposed; in truth, he did not suspect the prolonged gestation of Quaternary art, but contracted into one small division the documents of numerous phases. At the same time, Piette was right in attributing the flower of the art of engraving accompanied by contours of animal forms in relief to the second and third levels of the Magdalenian industry, but he had no idea that this development had been preceded by a long period in which engraving had been practised in a timid and more or less sporadic manner as a parietal art on the walls of the caverns as well as on bone and stone. It is also true that a considerable facility in sculpture preceded the art of engraving, but it was arrested in its progress while engraving slowly developed; in the early choice of subjects the sculptors of middle and late Aurignacian times showed a preference for the human form, while later, in Solutrean and early Magdalenian times, they inclined principally toward animal figures, so that sculpture was not suddenly eclipsed. The first engravings made with fine points of flint on stone are hardly less ancient than the first sculptures, and modestly co-exist beside them up to the moment where engraving, greatly multiplied, largely supplants sculpture. Finally, observes Breuil, it is one of the glories of Edouard Piette to have understood that the painted pebbles of Mas d'Azil represented the last prolongation of the dying Quaternary art. It is fortunate that the mantle of Piette fell upon a man of the artistic genius and appreciation of Breuil, to whom chiefly we owe our clear understanding of the chronological development of Upper Palæolithic art. In the accompanying table (p. 395) are assembled the results of the observations of Piette, Sautuola, Rivière, Cartailhac, Capitan, Breuil, and many others, largely in the order of sequence determined through the labors of Breuil. [Illustration: FIG. 196. Geographic distribution of the more important Palæolithic art stations of Dordogne, the Pyrenees, and the Cantabrian Mountains. After Breuil and Obermaier.]