SYSTEMS OF CLASSIFICATION 
123 
Systems of Classification. It is evident that in any 
written system of classification there must be a considerable 
element of artificial grouping. For practical purposes this 
matters little, and when once a system has been adopted for 
use it is well to adhere to it ; in large herbaria, for example, 
it is impossible to make frequent changes of arrangement to 
suit the progress of knowledge. Students should be trained 
in the system which offers the closest approach to the 
representation of the knowledge of the day. The system 
of Bentham and Hooker, proposed in their Genera Planta- 
rum , has remained the standard one in Britain for a long 
time, but on the Continent has long been superseded by 
the more natural system based on that of Brongniart. The 
latest exposition of this system is that of Engler in the 
Syllabus der Vorlesungen , and in Die natiir lichen Pflanzen- 
familien . We have therefore adopted this system for the 
flowering plants, but a key is also given to the system of 
Bentham and Hooker, so that those who prefer may use 
the latter. The natural orders given in Part II. are those 
of Engler, but their place in the other systems is always 
indicated. A third system, outlined below, is that of 
Eichler, adopted by Warming (, Systematic Botany , English 
ed.) ; it resembles that of Engler very closely. The ferns 
and their allies are classified according to the most recent 
views upon the subject. 
The Vegetable Kingdom is primarily divided into four 
groups : 
Thallophyta — Algae and Fungi. 
Bryophyta — Mosses and Liverworts. 
Pteridophyta or Vascular Cryptogams — Ferns, Lycopods, 
Equisetums, &c. 
Spermaphyta or Phanerogams — Seed-plants or so-called 
flowering plants (see Part II.). 
With the first two groups we are not concerned in this 
work. The classification of the Ferns, &c. is given in 
Part II. (art. Pteridophyta, &c.). We shall deal here with 
the last and highest group. 
The primary division of the Spermaphytes in all systems 
is into Gymnosperms and Angiosperms, but it is probable 
that this is very artificial — that the former are really three 
