rules, whereas the superior divisions, which are artificial, owe their existence to his disregard of them. De Candolle declared emphatically against the old notion, that the vegetable system answers to a linear series,—a notion which sprang from a misunderstanding of the saying, ‘Natura non facit saltus,’—and demonstrated its impossibility by examples; but he allowed himself to be too much influenced by the idea which had been thrown out by Linnaeus, and taken up by Giseke, Batsch, Bernardin de St. Pierre, L’Heritier, Du Petit-Thouars and others, that the vegetable kingdom might be compared as respects its grouping to a geographical map, in which the quarters of the globe answer to the classes, the kingdoms to the families, and so on. If the theory of descent is to a certain degree compatible with the idea of a linear sequence from the most imperfect to the highest forms of plants, it is quite incompatible with the above comparison; and systematic investigation, led astray from the right path, is in danger of ascribing the importance of real affinities to mere resemblances of habit, incidental analogies, by which a group of plants appears to be connected with five or six others. In exhibiting his system on paper De Candolle allowed the use of the linear sequence as a convenience, for here it was not, he said, a matter of any importance, since the true task of the science is to study the relations of symmetry in each family and the mutual relations of families to one another; yet in a linear presentation of the system for didactic purposes the sequence ought not to begin with the most simple plants, for these are the least known, but with the most highly developed. Thus De Candolle was the means of removing from the system the last trace of anything which harmonised with an ascending and uninterrupted development of forms. Resting on the doctrine of the constancy of species, and assuming that every group of relationship is founded on a plan of symmetry round which individual forms are grouped as crystals round their parent form, De Candolle was quite consistent in his views. The mode of representation came to prevail in the vegetable kingdom which De Candolle’s contemporary, Cuvier, an equally sturdy defender of the dogma of constancy, had introduced in the animal kingdom as the type-theory. Thus the most splendid results obtained by induction were united in the case of De Candolle with the barren dogma of the constancy of species, which, as Lange wittily remarks, comes direct from Noah’s ark, to form an intimate mixture of truth and error; nor did De Candolle’s many adherents succeed in unravelling the coil, though they removed the chief errors from his system and introduced many improvements. To these remarks may be appended a table of the main divisions of De Candolle’s system of 1819, which so far as it is presented in linear arrangement he calls expressly an artificial system. I. Vascular plants or plants with cotyledons.