Page:History of botany (Sachs; Garnsey).djvu/72

 necessarily be mixed up together ; the difficulty arises from our uncertainty as to the rules by which we should determine the resemblances of the genera. While there are two chief parts in plants, the root and the shoot, we cannot, as it seems, determine the genera and species from the likeness or unlikeness either of the one or of the other; for if we make a genus of those plants which have a round root, as the turnip, Aristolochia, Cyclamen, Arum, we separate generically things which agree together in a high degree, as rape and radish which agree with the turnip, and the long Aristolochia which agrees with the round, while at the same time we unite things most dissimilar, for the Cyclamen and the turnip are in every other respect of a quite different nature; the same is the case with divisions which rest merely on differences in the leaves and flowers.

In pursuing these reflections, which have the conception of species chiefly in view, he arrives at the following proposition: That according to the law of nature like always produces like, and that which is of the same species with itself.

All that Cesalpino says on systematic arrangement shows that he was perfectly clear in his own mind with regard to the distinction between a division on subjective grounds, and one that respects the inner nature of plants themselves, and that he accepted the latter as the only true one. He says, for instance, in the next chapter: 'We seek out similarities and dissimilarities of form, in which the essence ('substantia') of plants consists, but not of things which are merely accidents of them ('quae accidunt ipsis').' Medicinal virtues and other useful qualities are, he says, just such accidents. Here we see the path opened, along which all scientific arrangement must proceed, if it is to exhibit real natural affinities; but at the same time there is a warning already of the error which beset systematic botany up