created, whereas the species had come from these, as so many local deviations.
Who would deny that Linnæus 's work has facilitated the task of those that have come after him? Nevertheless, many species have been repeatedly subdivided. We must henceforth admit that when a species goes through a period of mutation, or has just gone through it, the number of elementary species that keep up their independent existence by the side of the parent species may be considerable, as we have seen with Œnothera. And it is easily understood that a tendency arises to denominate for convenience those numerous elementary species not by their own names, but by a collective name. This happens in most handbooks of systematic botany for well-known European plants, as, e. g., Draba verna, of which not less than 200 perfectly stable elementary species are known, Viola tricolor, etc.
Henceforth, however, we may no longer allow ourselves to be guided by opportunism. Systematic botany will have to take her watchword from physiology. De Vries has combined the qualities of the experimenter, who dares to look the physiological problem in the face, with those of the systematist, who observes and appreciates with uncommon sagacity the slightest shades of difference, and who with utmost delicacy of touch sifts and deals with species and races, mutations and variations.
The elementary species are stable. Selection calls forth different races within the limits of these species, but whenever selection ceases the races turn back to the parent form. The maximum deviation in these races is generally obtained after three or four generations of continuous selection; it takes about as many generations to bring back the parent form.
It is superfluous to say that many of these phenomena must be yet submitted to experimental investigation. De Vries has started this, and both in the domain of fluctuating variation and formation of races and in that of crossing and hybridizing he has already partly completed, partly only just commenced, elaborate experiments. Others besides himself have of late years analyzed the phenomenon of variety closely. The Cambridge zoologist, Bateson, has attempted to trace in his well-known work, 'Materials for the Study of Variation,' what it really is that variability offers towards the making of species, both in the most different species of animals and with respect to their divergent organs. He has, however, not seen his way out of the labyrinth, and although he came to the conclusion that it is not fluctuating variability which presides over the formation of species, but that a discontinuity must necessarily play a part, yet he, too, has committed himself to the assumption that the determination of the width of the fluctuations can furnish us with valuable data for understanding the gradual forma-