53 C. H. Ostenfeld. 
anum which A. Jordan has described (G. modestrum, G. villarsianum, 
etc.), the specimens collected in West Ireland on our excursion 
do not belong to them. Rouy and Foucaud (Fl. de France, vol. 4, 
1897, p. 95), admit under G. Robertianum (besides some forms 
of lower rank) only one “ forme” (i.e., sub-species), viz., G. pur- 
pureum Vill., and under this sub-species they have arranged a whole 
series of Jordan’s segregates. I think this arrangement is very 
good. The true G. Robertianum L. has more or less hairy cai pels 
which have rather few and distantly placed transverse ridges (otcen 
branching), mostly larger petals and red anthers; its leaves are 
several times pinnatifid. G. purpureum has glabrous carpels with 
many and densely placed transverse ridges (not branching), smaller 
petals and yellow anthers; its leaves are not so many times 
pinnatifid. 
According to the description a plant found in Cornwall (see 
Bot. Exchange Club Rep. for 1910), must belong to G. purpureum, 
but our Irish plants (from Clifden, Co. Galway, and Ballyvaghan, 
Co. Clare) do not; they have carpels of the true G. Robertianum, and 
must be referred to it. They have not the external habit of the 
typical G. Robertianum, and the flowers are considerably smaller, 
etc.; it is therefore probable that they represent a special variety, 
which has to be examined under cultivation and perhaps should have 
a name, but they must not be identified with G. Villarsianum Jord. 
or G. purpureum Vill. 
ROSACEZ2. 
Alchimilla acutidens, Buser, Lindb. fil. ampl. 
The record of this “ species” of Alch. vulgaris on Ben Lawers 
(alt. 400 to 800-1000 m.) is rather interesting from a phytogeo- 
graphical point of view. It is an alpine and northern species which 
according to H. Lindberg’s monograph (Die nordischen Alche- 
milla vulgaris -Formen und ihre Verbreitung, Helsingfors, 1909), 
is found in Russia, Finland, Sweden (southward to ca. 59’ N. Lat.), 
Norway, the Faröes and Iceland, besides the Alps. It is rather 
surprising that it was not known from the Scottish mountains 
before, although Dr. Lindberg had abundant material of Alchi- 
milla from the British Isles for examination. Now, that we have 
found it on Ben Lawers,—and no doubt it will be found elsewhere 
in the Scottish Highlands,— a gap in its area of distribution has 
been filled up. 
The “ species ” is very near the common A. alpestris Schmidt, 
from which it differs in the following points :— 
The flowering stems sparsely covered with adpressed hairs in 
the lower part up to the second (or third) branch ; the leaf-stalks also 
bear adpressed hairs, as do the nerves of the leaves on the under- 
side and generally the outermost part of the basal lobes; leaves 
orbicular or reniform with 9(-11) lobes; teeth uniform, acute ; 
apical tooth as long as and uniform with the others. Apart from 
the distribution of hairs here described the plant is glabrous (with 
the exception of the hairs on the tips of the leaf-teeth). A. 
alpestris bears still fewer hairs; only the leaf-nerves of the underside 
are hairy on their distal halves, and the stems and leaf-stalks are 
very slightly hairy; the teeth of the leaves are not so uniform nor 
acute, and apical tooth is smaller than the others. 
