NECESSARY PRELIMINARY KNOWLEDGE.

For the generic character the set of features that defines a group, some single property or mark common to all already known species and unique to them alone must be chosen, wherever it can be found. That mark is most conveniently taken from the Flower or Fruit in plants that have these parts, because these parts usually vary less within the same genus than others, and they are notable, conspicuous, and easily observable due to their color, shape, and other accidental features qualities like color or size that do not change what a thing is, but help describe it. If, however, some plant should occur in the future which agrees with other species of a genus in its other parts and features, but lacks this single mark which we have adopted as the Characteristic; despite this, we establish that it should be referred to that genus with which it has more common attributes.

In my Letter to Mr. Rivinus Augustus Quirinus Rivinus (1652–1723), a German botanist who argued for a simpler classification system based solely on the number of flower petals. and my Dissertation on the Various Methods of Plants, I decided that in those genera where such a notable property—common to all species contained within them and unique to them alone—is lacking, a "complex of several common attributes" should be taken as the Characteristic. This complex would consist of features that all together would belong to all the species contained under those genera, and to no others besides. But when I consider how—I do not say difficult—but impossible it is to assign a certain number of such accidental features, unless one knows every species of plant growing throughout the whole world (For how can one be certain that among those which still lie hidden unobserved, some will not be found which agree in their whole habit the general appearance or growth form of a plant and other marks with a certain described species, yet lack one or another mark from the assigned Characteristics?), since this is so, I judge it sufficient that all the assigned attributes belong to most species of a genus, and most of them to all. Nor should any species be excluded from fellowship or participation in a genus because it lacks one or another mark from those which [make up] the characteristi-