causes, by which the original symmetry of a class is changed and may even be utterly obscured, are then fully illustrated by examples. In respect to abortion he distinguishes that which is produced by internal causes from that which is due to accidental and external ones; he refers especially to the abortion of two loculaments in the fruit of the horse-chestnut and the oak, to the suppression of the terminal bud in some shrubs by the adjoining axillary buds, and to the fact that all organs of plants may become abortive in a similar manner; for instance, the sexual organs disappear entirely in the disk-flowers of Viburnum Opulus, and one of the two sexes in the flower of Lychnis dioica. He goes on to answer the question, how it is possible to discover the symmetry in such cases; one method he finds supplied by monstrosities, among which there are even some that may be regarded as a return to the original symmetry, the cases known as peloria. Analogy or ‘induction’ is, he says, less certain, but of much more extensive application; this is founded exclusively on the knowledge of the relative position of organs. Armed with this, we find that the flower of Albuca, which corresponds to a flower of Liliaceae in everything except in having only three stamens, is to be considered one of the Liliaceae, because it has three filaments placed between the three stamens exactly in the position of the three other stamens in the Liliaceae; it must be concluded therefore that they are abortive stamens. Similar conclusions from analogy must be carried from species to species, from organ to organ, and the great systematists have in fact done so. In certain cases abortion is produced by defect, in others by excess of nourishment, of which he gives examples. An important sentence occurs in this place; everything in nature, he says, leads us to believe that all organisms in their inner nature are regular, and that different forms of abortion differently combined are the cause of all irregularity; from this point of view the smallest irregularities are important, because they lead us to expect greater ones in nearly allied plants; and wherever in a given system of organisation there are inequalities between organs of the same name, the inequality will possibly reach a maximum, that is, end by annihilating the smallest part. Thus in the Labiatae with two stamens, it is the two which in other cases also are the smaller, which are here completely aborted. When in Crassulaceae there are twice as many stamens as petals, those that alternate with the petals are larger and earlier developed, and we may therefore expect that those which are opposite the petals may become abortive; and therefore we may place a genus like Sedum, in which the latter are sometimes wanting, with Crassulaceae; but we could not do so, if we found only the stamens that are superposed upon the petals. It occurs sometimes, he continues, that an organ is prevented