Page:History of botany (Sachs; Garnsey).djvu/135

 point, because it expresses the idea, that the different genera of such a group are to some extent regarded as forms derived from the one selected to supply the name. Many of Linnaeus' orders do in fact indicate cycles of natural affinity, though single genera are not unfrequently found to occupy a false position; at all events, Linnaeus' fragment is much the most natural system proposed up to 1738, or even to 1751. It is distinguished from Kaspar Bauhin's enumeration in this, that its groups do not run into one another, but are defined by strict boundaries and fixed by names.

The Linnaean list is distinctly marked by the endeavour to make first the Monocotyledons, then the Dicotyledons, and finally the Cryptogams follow one another; that the old division into trees and herbs already rejected by Jung and Bachmann, but still maintained by Tournefort and Ray, disappears in Linnaeus' natural system will be taken for granted after what has been already said of it, and from this time forward this ancient mistake is banished for ever.

In 's arrangement of 1759 we find some improvements in the naming, the grouping, and the succession, but at the same time some striking offences against natural affinity. He published no theoretical remarks on the system, but gave expression to his views on relations of affinity in the vegetable kingdom in laying out the plants in the royal garden of Trianon, and in the garden-catalogue. His nephew published his uncle's enumeration in the year 1789 in his 'Genera Plantarum,' affixing the date of 1759 given above. The difference between it and the Linnaean fragment does not seem