The Principal Conventions in Egyptian Sculpture. 319 just to assert that Egyptian art had two difterent and successive canons. The question as to whether the Egyptians ever adopted a unit of measurement in their rendering of the human figure or not, is different. Wilkinson and Lepsius thought they had discovered such a unit in the length of the foot, Prisse and Ch. Blanc in that of the medius. There is nothing in the texts to support either theory, and an examination of the monuments themselves shows that sometimes one, sometimes the other of the two units, is most in accordance with their measurements. Between the Ancient Empire and the New proportions differed so greatly that it is impossible to refer them to one unit. Among the works of a single period we find some that may be divided exactly by one of the two ; others w^hich have a fraction too much or too little. It has not yet been proved, therefore, that the Egyptians ever adopted such a rigorous system as that attributed to them. Like all races that have greatly practised design, they established certain rela- tions between one part of their figures and another, relations which gradually became more constant as the national art lost its freedom and vitality ; and they arrived at last at the mechanical reproduction of a single figure without troubling themselves to calculate how many lengths of the head, the nose, the foot, or the medius, it might contain. Their eyes were their compasses, and they worked — at least under the New Empire and during the Grseco-Roman period — from models which represented the expe- rience of the past. It is therefore unnecessary to search for an explanation of the uniformity which characterises their w"orks in the following of a rigid mathematical system ; we must be content to see in it the natural result of an artistic education into which, as the centuries succeeded one another, the imitation of previous types, and the application of traditional recipes entered more and more. As for the designs traced within lines which cross each other at regular intervals, they can be nothing but drawings squared for transferring purposes. Squaring is the usual process employed by artists when they wish to repeat a figure in different dimensions from those of the original. Having divided the latter by hori- zontal and perpendicular lines cutting each other at regular intervals, they go through the same operation upon the blank surface to which the figure is to be transferred, makino- the lines
