19 
rupicola .” “ Sorbus meridionalis ,” Gass., ia Strobbi’s “ Flora 
Nebrodensis,” appears intermediate between P. grceca and P. eu-aria , 
but differs from both in having the central tooth of each lobe of the 
leaf (that into which the side veins run) much larger and more 
acuminate-acute than the others, as in P. latifolia , to which it also 
approaches in the subcoriaceous texture of the leaves ; the veins are 
from 6 to 10 in number; the margins of the leaves have minute 
deltoid-acute lobes, deepest in the uppermost fourth of the leaf, and 
the under surface of the leaf is snow-white. 
3. Pyrus latifolia. Whether or no we consider Pgr us rupicola 
sufficiently distinct from P. eu-aria to require a separate specific name, 
there is a very general feeling that the plant which I believe ought to 
be called P. latifolia is something more than a variety of P. Aria 
By British Botanists it has till within a few years been assumed to be 
Sorbus scandica of Fries., the Cratcegus Aria var. a scandica of Linn., 
Amoen., and the C. Aria, 8 suecica, Linn. Sp. PI. So far back as 
1851, however, Prof. C. C. Babington in the “Botanical Gazette,” 
vol. iii., p. 34, in speaking of the (English) Sorbus scandica, at that 
time the only one recognised, says, “Nearly allied to and, as I think, 
not distinct from this, is the S. latifolia, Pers.” In this, as far as 
the English P. scandica goes, I quite agree with him. It was not 
until the year 1869, five years after the description of P. scandica for 
“ Engl. Bot.,” ed. iii., was written, that it dawned upon my mind that 
the English scandica was not the same as the Scandinavian. See the 
report of Bot. Exchange Club, 1869, p. 11, in which I mention that 
the Devonshire specimens show an approximation towards S. latifolia, 
but at that time I fell into two errors from having only imperfect 
Continental specimens. I began to think that the English scandioa 
was S. Mougeoti. The other error will be mentioned below under 
Pyrus fcnnica. This error was confirmed by receiving a Continental 
specimen of P. latifolia, as I now believe labelled Mougeoti, which 
led me to think that P. Mougeoti and latifolia were forms of one plant. 
More recently the receipt of further specimens of S. Mougeoti and 
latifolia, and the perusal of Mons. Grenier’s “ Flore de la Chaine 
Jurassique” showed me that S. Mougeoti and latifolia were not 
identical, and that the English scandica was the Continental latifolia. 
The most recent contribution to my knowledge has been five specimens 
named Sorbus latifolia , in Dr. Reichenbach’s collection, which Mr. 
Watson kindly placed at my disposal. These five specimens include 
nearly the whole range of forms of our British plant which have come 
under my observation, so that now I have no doubt that to the plant 
now under consideration the name latifolia ought to be applied. 
b 2 
