Tlie Groirth- Forms of Natal Plants. 
619 
Acanthaceae. Among the most troublesome are those forms (the Labiate 
type of undershrub of Warming) where considerable parts of the flowering 
shoots die after blossoming. Since one is naturally acquainted more with the 
appearance of the plant at the time when it flowers, one is often inclined to 
put it down as Nanophanerophyte when it ought to be classed as Chamae- 
phyte. 
The total number of ISTanophanerophytes is cir. 430. They form a 
somewhat heterogeneous class, much more so than any of the preceding 
classes. This being the case, some further subdivision is possible, and such 
subdivision can follow various lines. 
As is well known to all botanists, certain large families of flowering 
plants have a striking similarity in their vegetative parts, so much so that 
it is often possible to tell at a glance to what family a plant belongs, without 
examining the flowers. Taking advantage of this fact, we might pick out 
certain types of growth-form from this class and name each after the natural 
order, of which it is most characteristic. 
Such a simple scheme of subdivision is one that readily suggests itself. 
The earliest writer on growth-forms, Kumholdt (1805), adopted it. He 
treated in detail the forms of the palm, banana, malvaceous and bombaceous 
plants, mimosa, heath, cactus, orchid, casuarina, conifer, pothos, liane, aloe, 
grass, fern, lily, willow, myrtle, melastomaceous plant, laurel. 
Though it is interesting enough to recognise such " family resemblances " 
in growth-forms, yet if we attempt to push it very far it has really very 
little scientific value, as the scheme is not based on ecological facts. The 
members of no family of plants all belong to the same growth-form, and 
after we have removed the forms belonging to each of the types mentioned 
above, or to any similar collection of types, we are left with a still more 
heterogeneous collection, to which we have been unable to assign places in 
any of the types. Not only, therefore, is such a scheme not based on 
ecological facts, but it is not exhaustive. It is simply a somewhat super- 
ficial recognition of certain resemblances in the vegetative organs of plants 
without any reference to any underlying principle or explanation. 
The attempt to apply it generally, as Humboldt did, to the v.diole of the 
flowering plants must necessarily fail. At the same time, there is much that 
is attractive about the system of nomenclature. It can be made quite 
scientific if we adopt some definite ecological factor as a basis for the scheme, 
and it can be applied successfully provided we are dealing with one definite 
ecological group, such as the Nanophanerophytes . In other words, though 
it cannot successfully be applied as a basis for the main ecological divisions, 
yet it can be used in minor subdivisions. There is little likelihood of any 
confusion arising from introducing systematic names into ecology. No 
one, for instance, is likely to suppose that when we speak of " Heath 
form " or " Acanthaceous form " we restrict ourselves to Ericaceae and 
