17 
ago accorded specific rank to three of these forms; one of the three he 
referred to another genus. The synonymy cited shows that similar 
if varying views had been held by other writers. The treatment 
by Miiller, whose careful monograph of 1866 has not since then 
been critically examined, involved the recognition of four of these 
five forms, three of them as varieties of A. decumbens, the last as 
a distinct species, A. grandidentata. 
More closely examined, however, this last form, though at first 
sight apparently the most distinctive of all, proves in reality to be 
undeserving of separate recognition. It differs from Urtica capensis, 
as described by the younger Linnaeus in 1781 and from Tragia 
villosa, as described by Thunberg in 1794, which is the same as A. 
Kraussiana, described by Meisner in 1845, only in having the female 
inflorescences aggregated in a terminal spike in place of being 
e e 
different facies thus imparted to these two plants, A. pieutlidedt Wh 
is merely a condition of A. Kraussiana, Meisn. (= A. decumbens, 
var. villosa, Mill. Arg.). 
The distinction between the typical A. decumbens of Thunherg 
and the form which that author described as A. cordata is hardly 
more tangible. In the original specimens of A. decumbens the 
leaves are all reddish-hoary underneath and are all truncate or only 
slightly cordate at the base, while in the original specimens of 4. 
cordata the leaves are all olive-hoary beneath and are nearly all 
distinctly cordate at the base. But it was not owing to these real 
differences that the two forms were separated as species by Thun- 
rg; the main character relied on for their separation was that 
A, decumbens is herbaceous while A. cordata is shrubby. Thunberg’s 
belief we know now to be without foundation; we now know too that 
while the main branches have leaves that conform with those of the 
original A. cordata, the secondary branches have leaves that agree 
with those of the original A. decumbens. This was fully appreciated 
by E. Meyer in dealing with Drége’s specimens, some of which, and 
some also of Zeyher’s, show both forms on the same branch; the name 
A, discolor, suggested by E, Meyer, happily indicates the peculiar 
difference in the colour of the underside of the cordate and the less 
instead of having the nerves very shortly puberulous. Now it is found 
that even this distinction breaks down, since in certain specimens 
collected by Marloth we find leaves characteristic of A. cordata and 
leaves characteristic of A. lamiifolia in plants which have grown 
side by side. The true position of the form Jamiifolia is midway 
between the forms cordata and villosa and the true significance of 
the form does not lie in its differences from these two, but in its 
testimony that they themselves do not really differ from each other. 
Taving regard, however, to the extreme diversity of view which 
has hitherto prevailed, and to the convenience from the biblio- 
graphical standpoint which the division of the species involves, it 
