231 
left undone if the agricultural activities of the Department 
are to have full scope. And this has been done without any 
loss of their value to the pure scientific study of botany. In 
fact it has largely increased it, for laboratories and other fa- 
cilities have been provided for the departmental work of the 
gardens, and have proved of great value for the prosecution of 
pure scientific research, which might otherwise have hoped in 
vain to get them provided. It will be well, therefore, to study 
in more detail what are the functions of a modern garden. 
Agriculture in the long run depends upon progress in the 
subsidiary sciences of botany, zoology, chemistry and others, 
and we must therefore do all we can to promote advance in 
these. Now progress in science, and indeed in many other 
things, is most favoured by exact and systematic research, and - 
this consequently is the chief aim which should be set before 
a modern botanic garden. : 
The work of the older botanic gardens was based simply on 
systematic botany, which indeed is the oldest branch of the 
subject, and the development of a modern botanic garden may 
be said to reflect the growth of botany in general. To the 
older study of the systematic botany of the flowering plants 
has been added the study of the fungi, of vegetable physiology, 
of plant breeding and evolution, and of many other lines 
of work. 
However much it may attend to new departments of research, 
a botanic garden must not neglect the old, and the completion 
of the flora of the country, the formation of a good herbarium 
and museum of its products, and the establishment of a great 
Collection of at least two living specimens of every one of 
the plants known in it, must not be forgotten. This line of 
work has not suffered at Buitenzorg, and is now well on the 
way to completion. / 
But such work simply affords a sure basis upon which to 
proceed with further research. The vegetable physiologist, the 
chemist, and others may go on to study the local plants in 
detail, to find out whether more of them may not prove of 
