Tournefort's system is thoroughly artificial, if possible, more artificial than that of Bachmann, and certainly inferior to Ray's. If we meet with single groups that are really natural, it is simply because in some families the genera so agree together in all their marks, that they necessarily remain united, whatever mark we select for the systematic purpose. We do not find in Tournefort the distinction between Phanerogams and Cryptogams already established by Ray, nor the division of woody plants and herbs into Monocotyledons and Dicotyledons; if his chief work, to which we confine ourselves here, the 'Institutiones rei herbariae,' did not bear the date of 1700, we might conclude that it was written before the 'Historia Plantarum' of Ray, and the chief work of Bachmann. Yet it has one merit of a purely formal kind; it is pervaded by a rigorous spirit of system; every class is divided into sections, these into genera, and these again into species; figures of the leaves and of the parts of the flower, very beautifully engraved on copper-plate and filling a whole volume, are perspicuously arranged; the whole work therefore is easy to consult and understand. But to form an idea of the confusion as regards natural affinities that reigns in his system, we need only examine the first three sections of his first class, when we shall find Atropa and Mandragora together in the first section, Polygonatum and Ruscus in the second, Cerinthe, Gentiana, Soldanella, Euphorbia, and Oxalis in the third. The handiness of the book, the little interest taken by most of the botanists of the time in the question of natural relationship, and the continually increasing eagerness for a knowledge of individual plants, are evidently the reasons why Tournefort gained over to his side most of the botanists not only of France, but also of England, Italy, and Germany; and why later attempts in systematic botany during the first thirty or forty years of the 18th century were almost exclusively founded on his system, as they were afterwards on the sexual system of Linnaeus. Boerhaave, among others, proposed a system in 1710, which may be regarded as a combination of
