Page:History of botany (Sachs; Garnsey).djvu/90

 and monocotyledonous embryo, yet he may claim the great merit of having founded the natural system in part upon this difference in the formation of the embryo. He displays more conspicuously than any systematist before Jussieu the power of perceiving the larger groups of relationship in the vegetable kingdom, and of defining them by certain marks; these marks moreover he determines not on a priori grounds, but from acknowledged affinities; but it is only in the great divisions of his system that he is thus true to the right course; in the details he commits many and grievous offences against his own method, as we shall see below when we come to an enumeration of his classes. Modern writers have often attributed to Ray the merit of having first taught the transmutation of species, and of being thus one of the founders of the theory of descent. Let us see how much truth there is in this assertion. Though plants, says Ray, which spring from the same seed and produce their species again through seed, belong to the same species, yet cases may occur in which the specific character is not perpetual and infallible. Seeds may sometimes degenerate and produce plants specifically distinct from the mother-plant, though this may not often happen, and so there would be a transmutation of species, as experience teaches. It is true that he considered the statements of various writers, that Triticum may change into Lolium, Sisymbrium into Mentha, Zea into Triticum, etc., to be very doubtful, yet there were, he thought, other cases which were well ascertained; it was in evidence in a court of law that a gardener in London had sold cauliflower seed which had produced only common cabbage. It is to be observed, he says, that such transmutations only occur between nearly allied species and such as belong to the same genus, and some perhaps would not allow that such plants are specifically distinct. These words, especially when judged by Ray's general views, appear only to express the opinion that certain inconsiderable variations are possible within a narrow circle of affinity, especially in