It is curious that Cesalpino, who, as we have pointed out, had arrived at the very important principle that the seed and fruit characters were of major value in classification, yet put forward a system which was distinctly inferior to that of Gaspard Bauhin, although the latter appears to have been guided by no such general principles. Probably the reason for this is to be sought in the fact that no system of classification can represent natural affinities, unless it takes into account the nature of the plant as a whole. It is true that, compared with the characters of the reproductive organs, the leaf-form and habit, owing to their plasticity, have to be used with great discretion as systematic criteria, but, nevertheless, no system of classification can afford to ignore them entirely. Cesalpino based his scheme too exclusively upon seed characters, to the neglect even of the structure of the flower, and, curiously enough, although he laid so much stress upon the nature of the seed, he did not grasp the fundamental distinction between the embryos of the Monocotyledons and the Dicotyledons, due to the possession of one, and two seed-leaves respectively. The chief drawback of his scheme, however, was his failure to realise that living organisms are too complex to fall into a classification based on any one feature, important as that feature may prove to be when used in conjunction with other characters. Those herbalists, on the other hand, who attacked the problem of the classification of plants without any preconceived, academic theory, depended, one might almost say, on the glimmerings of common sense for the recognition of affinities. This was no doubt a dim and fitful illumination, but it was at least less partial than the narrow, lime-light beam of a rigid theory. CHAPTER VII THE EVOLUTION OF THE ART OF BOTANICAL ILLUSTRATION In the art of botanical illustration, evolution was by no means a simple and straightforward process. We do not find, in Europe, a steady advance from early illustrations of poor quality to later ones of a finer character. On the contrary, among the earliest extant drawings, of a definitely botanical intention, we meet with wonderfully good figures, free from such features as would be now generally regarded as archaic. The famous Vienna manuscript of Dioscorides (see pp. 8 and 85) is a remarkable example of the excellence of some of the very early work. It dates back to the end of the fifth, or the beginning of the sixth century of the Christian era. It is illustrated with brush drawings on a large scale, which in many cases are notably naturalistic, and often quite modern in appearance (Plates I, II, XV). The general habit of the plant is admirably expressed, and occasionally, as in the case of the Bean (Plate XV), the characters of the flowers and seed-vessels are well indicated. In this drawing, also, the leaves are effectively foreshortened.